CN111126393A - Vehicle appearance refitting judgment method and device, computer equipment and storage medium - Google Patents
Vehicle appearance refitting judgment method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN111126393A CN111126393A CN201911354664.7A CN201911354664A CN111126393A CN 111126393 A CN111126393 A CN 111126393A CN 201911354664 A CN201911354664 A CN 201911354664A CN 111126393 A CN111126393 A CN 111126393A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- standard
- component
- image
- label
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000012545 processing Methods 0.000 claims abstract description 70
- 230000000877 morphologic effect Effects 0.000 claims abstract description 48
- 238000010586 diagram Methods 0.000 claims abstract description 33
- 230000011218 segmentation Effects 0.000 claims description 27
- 239000003550 marker Substances 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 239000000284 extract Substances 0.000 claims description 13
- 230000008859 change Effects 0.000 claims description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000012986 modification Methods 0.000 abstract description 12
- 230000004048 modification Effects 0.000 abstract description 12
- 238000007689 inspection Methods 0.000 description 10
- 238000001514 detection method Methods 0.000 description 9
- 239000003086 colorant Substances 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 238000013136 deep learning model Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012857 repacking Methods 0.000 description 2
- 238000009423 ventilation Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 239000007795 chemical reaction product Substances 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The application provides a vehicle appearance modification judgment method and device, computer equipment and a storage medium, and a vehicle image and a vehicle standard image are obtained; segmenting the vehicle image and the vehicle standard image to obtain a vehicle image mask image and a vehicle standard mask image; extracting a component sub-mask diagram of the target component from the vehicle mask diagram to generate a component label diagram, and extracting a component standard sub-mask diagram of the target component from the vehicle standard mask diagram to generate a component standard label diagram; performing morphological processing on the component label image to obtain component comparison information, and performing morphological processing on the component standard label image to obtain component standard comparison information; comparing the information to be compared of the component with the component standard comparison information to obtain a comparison result, and judging whether the comparison result is within a set error range; and analyzing the comparison result, and judging whether the vehicle appearance is modified or not. According to the method, whether the appearance of the vehicle is modified or not can be judged quickly, manual judgment is replaced, and the working efficiency is improved.
Description
Technical Field
The patent relates to the field of artificial intelligence image processing, in particular to a method and a device for judging vehicle appearance modification, computer equipment and a storage medium.
Background
With the continuous development of social economy and the continuous improvement of the living standard of people, the holding quantity of vehicles is continuously increased, the annual inspection of the vehicles is required every year, and therefore the workload of the annual inspection of the vehicles is rapidly increased. The detection of vehicle outward appearance repacking in traditional vehicle annual check is mainly through artifical the detection, to whether qualified, the condition of whether repacking of vehicle outward appearance, general inspection personnel all judge according to the experience, and the subjectivity is stronger, and the measurement standard of outward appearance judgement is more in addition, and some circumstances are more complicated, and artifical the inspection has very big easy tired, drawback such as neglected easily, and artifical check-up cost is higher simultaneously. Therefore, how to accurately and quickly judge the appearance of the vehicle, particularly the appearance of a large vehicle, is an urgent problem to be solved.
Disclosure of Invention
The application provides a vehicle appearance refitting judgment method and device, computer equipment and a storage medium, overcomes the defects of manual inspection in the prior art, realizes the detection of the vehicle appearance by using the artificial intelligence of a deep learning model, replaces manual detection, provides the working efficiency and unifies the measurement standard.
In order to achieve the above object, in one aspect, the present application provides a vehicle refitting determination method, including: acquiring a vehicle image and a vehicle standard image corresponding to the vehicle image; segmenting the vehicle image to obtain a vehicle mask image, and segmenting the standard image to obtain a vehicle standard mask image; extracting a component sub-mask diagram of a target component from the vehicle mask diagram to generate a component label diagram, and extracting a component standard sub-mask diagram of the target component from the vehicle standard mask diagram to generate a component standard label diagram; performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information; comparing the information to be compared with the standard comparison information of the component to obtain a comparison result, and judging whether the comparison result is within a set error range; and analyzing the comparison result, and judging whether the vehicle appearance is modified or not.
Further, performing morphological processing on the component tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component comparison information, performing morphological processing on the standard tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component standard comparison information, and including: setting a comparison reference object, wherein the reference object is a mark post, and acquiring the actual physical size of the mark post; calculating the ratio of the height of the marker post in the marker post label graph to the actual physical size to obtain a first ratio, and calculating the ratio of the height of the marker post in the marker post standard label to the actual physical size to obtain a second ratio; acquiring a first size of the target component in the component label diagram and a second size of the target component in the component standard label diagram; and calculating the information to be compared of the component according to the first size and the first proportion, and calculating the standard comparison information of the component according to the second size and the second proportion.
Further, performing morphological processing on the component tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component comparison information, performing morphological processing on the standard tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component standard comparison information, and including: performing morphological processing on a target area of the component label map, removing noise points, and combining the same targets in the component label map to obtain the target component interesting area label map only containing the target component interesting area; and performing morphological processing on the component standard label map, removing noise points, and combining the same targets in the component standard label map to obtain the target component interested area standard label map only containing the target component interested area.
Further, the vehicle appearance refitting judgment method further includes: extracting a first region of interest from the target component region of interest label map, and calculating the first size of the target component according to the first region of interest; and extracting a second region of interest from the standard label map of the region of interest of the target component, and calculating the second size of the target component according to the second region of interest.
Further, performing morphological processing on the component tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component comparison information, performing morphological processing on the standard tag map, extracting the outline, the relative position, the overlapping degree and the scale information of the target component to obtain component standard comparison information, and including: the target component is a vehicle tail sign board; extracting a tail sign board area from the tail sign board label map, carrying out binarization and morphological processing on the tail sign board area, counting the number of times of color change in the area and the color level of a pixel value, and obtaining tail sign board comparison information; and extracting a vehicle tail sign board area from the vehicle tail sign board standard label map, performing binarization and morphological processing on the vehicle tail sign board area, counting the number of times of color change in the area and the color level of a pixel value, and obtaining vehicle tail sign board standard comparison information.
Further, analyzing the comparison result to determine whether there is a modification in the vehicle appearance, including: the target component is at least one target component, each target component corresponds to one comparison result, and if the target components are multiple, multiple comparison results are obtained; if any comparison result does not meet the preset condition, judging that the vehicle is modified, and judging that the final result is false; and if all the results meet the preset conditions, judging that the vehicle is not modified, and obtaining a true final result.
Further, a semantic segmentation model is adopted to segment the vehicle image to obtain the vehicle mask image, and the standard image is segmented to obtain the vehicle standard mask image; the semantic segmentation model training process is as follows: acquiring a vehicle image training set; marking a vehicle outline in the vehicle image, setting the values of all pixel points outside the vehicle outline to be 0, and extracting a vehicle area image; marking the outline of the 22 types of target components in the vehicle area image, setting corresponding label values for all pixel points in the outline of the 22 types of target components, and obtaining a target component pixel label image after marking; graying the labeled image of the target component pixel, forcibly modifying the value of the edge pixel point by using a nearest neighbor algorithm, generating a label pair with the original image of the vehicle image, and putting the label pair into a semantic segmentation model for training; in the training process, the foreground and background areas of the whole vehicle need to be preferentially learned, then the super-parameters and the category weights of the network are finely adjusted, and finally the semantic segmentation model is obtained.
In another aspect, the present application further provides a vehicle exterior refitting determination device, including: the acquisition module acquires a vehicle image and a vehicle standard image corresponding to the vehicle image; the segmentation module is used for segmenting the vehicle image to obtain a vehicle mask image and segmenting the standard image to obtain a vehicle standard mask image; the tag map generation module extracts a component sub-mask map of a target component from the vehicle mask map to generate a component tag map, extracts a component standard sub-mask map of the target component from the vehicle standard mask map to generate a component standard tag map; the image processing module is used for performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information; the comparison module compares the information to be compared of the component with the component standard comparison information to obtain a comparison result, and judges whether the comparison result is within a set error range; and the judging module analyzes the comparison result and judges whether the vehicle appearance is modified or not.
In yet another aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.
In yet another aspect, the present application also provides a computer readable storage medium having a computer program stored thereon, which when executed by a processor, performs the steps of the above method.
Compared with the prior art, the vehicle appearance refitting judgment method and device, the computer equipment and the storage medium acquire the vehicle image and the corresponding vehicle standard image; segmenting the vehicle image to obtain a vehicle mask image, and segmenting the standard image to obtain a vehicle standard mask image; extracting a component sub-mask diagram of a target component from the vehicle mask diagram to generate a component label diagram, and extracting a component standard sub-mask diagram of the target component from the vehicle standard mask diagram to generate a component standard label diagram; performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information; comparing the information to be compared with the standard comparison information of the component to obtain a comparison result, and judging whether the comparison result is within a set error range; and analyzing the comparison result, and judging whether the vehicle appearance is modified or not. The invention mainly applies a vehicle appearance refitting judgment method based on deep learning, realizes automatic verification of vehicle appearance, and can transmit failed verification images and reasons back to a server for storage and evidence collection, thereby saving manpower and ensuring the justice and the disclosure of verification work. The method effectively improves the working efficiency, can properly replace the fussy and non-standard manual inspection, and improves the accuracy of detection.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:
fig. 1 is a flowchart of a method for determining refitting of a vehicle appearance according to an embodiment of the present application;
fig. 2 is a flowchart of another vehicle appearance refitting determination method provided in the embodiment of the present application;
fig. 3 is a flowchart of another method for determining refitting of a vehicle appearance according to an embodiment of the present application;
fig. 4 is a flowchart of another method for determining refitting of a vehicle appearance according to an embodiment of the present application;
fig. 5 is a flowchart of another method for determining refitting of a vehicle appearance according to an embodiment of the present application;
fig. 6 is a flowchart of another method for determining refitting of a vehicle appearance according to an embodiment of the present application;
FIG. 7 is a flowchart of a segmentation model training method provided in an embodiment of the present application;
fig. 8 is a block diagram of a vehicle appearance modification determination apparatus according to an embodiment of the present application;
fig. 9 is a block diagram of another vehicle appearance modification determination apparatus provided in the embodiment of the present application;
fig. 10 is a block diagram of another vehicle refitting determination device provided in the embodiment of the present application;
the same or similar reference numbers in the drawings identify the same or similar elements.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In a typical configuration of the present application, the terminal, the device serving the network, and the trusted party each include one or more processors (e.g., Central Processing Units (CPUs)), input/output interfaces, network interfaces, and memory. The Memory may include volatile Memory in a computer readable medium, Random Access Memory (RAM), and/or nonvolatile Memory such as Read Only Memory (ROM) or flash Memory (flash RAM). Memory is an example of a computer-readable medium.
The vehicle appearance refitting judgment method provided by the application can be applied to the following application environments. The application environment comprises a terminal, a server and an image acquisition device. A terminal may refer to an electronic device with greater data storage and computing capabilities. Wherein, the terminal communicates with the server through the network. The image acquisition device can be in the terminal or can be a separate device. The terminal can be provided with a deep learning model which is trained. Specifically, the vehicle image is acquired through the image acquisition device, the vehicle standard image can be stored in the server, and the terminal acquires the vehicle image and the vehicle standard image. And the terminal adopts a deep learning model to carry out semantic segmentation on the vehicle image and the vehicle standard image to obtain a mask image. And the terminal generates a label graph through the mask graph and performs morphological processing on the label graph to obtain the information to be compared of the component and the component standard comparison information. And the terminal calculates a comparison result according to the obtained comparison information, analyzes the result by comparison and judges whether the vehicle appearance is modified or not.
In other embodiments, the vehicle appearance refitting judgment method provided by the application can also be applied to a terminal side and a server side, the image acquisition device acquires an image to be detected, the image to be detected is sent to the server through the terminal in a network connection mode and the like, and then the server detects and judges the vehicle appearance according to the image to be detected. The terminal can be, but is not limited to, various portable mobile devices, and the server can be a live server or a remote server.
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. It should be noted that, in the vehicle appearance determination method provided by the present application, the execution subject in fig. 1 to fig. 10 is a computer device, wherein the execution subject may also be a vehicle appearance determination apparatus, and the apparatus may be implemented as part or all of vehicle appearance detection by software, hardware, or a combination of software and hardware.
In one embodiment, as shown in fig. 1, the present application provides a vehicle refitting determination method, the method comprising:
and step S100, acquiring a vehicle image and a vehicle standard image corresponding to the vehicle image.
In one embodiment, the vehicle image can be obtained from a terminal such as an image acquisition device or a server of an institution such as a vehicle inspection station, and the appearance of the vehicle in the vehicle image is clear and clean, so that system and model identification is facilitated. The vehicle standard image can be obtained from a server side of the vehicle inspection station, the vehicle standard image corresponds to the vehicle image, and the vehicle standard image is a standard non-refitted appearance of the vehicle appearance in the vehicle image and is used for judging whether the vehicle appearance in the vehicle image is refitted or not.
S200, segmenting the vehicle image to obtain a vehicle mask image, and segmenting the standard image to obtain a vehicle standard mask image.
In one embodiment, the vehicle image is segmented to obtain a vehicle mask image, namely a mask image, and the standard image is segmented to obtain a vehicle standard mask image, namely a mask image. The image semantic segmentation model can be adopted to carry out on the vehicle image and the vehicle standard image, and the background pixel 0 of the mask image is segmented.
Further, the image is segmented to obtain the vehicle mask image, and a segmentation method of a semantic segmentation model of the image, such as a PSPNet model, a REFINENet model, and the like, may be adopted, which is not limited herein, and the semantic segmentation model is a conventional technical means in the art and is not described herein again.
S300, extracting a component sub-mask image of the target component from the vehicle mask image to generate a component label image, and extracting a component standard sub-mask image of the target component from the vehicle standard mask image to generate a component standard label image.
In one embodiment, for the judgment of the vehicle appearance, the appearance component to be detected can be set according to the actual detection requirement, and one target component or a plurality of target components can be selected. And for each vehicle target component, a set label value exists, a component sub-mask map of the target component is extracted from the vehicle mask map, and a component label map is generated by combining the label value of the target component. Similarly, the same operation and processing are carried out on the vehicle standard mask map, a component standard sub-mask map of the target component is extracted from the vehicle standard mask map, and the component standard label map is generated by combining the label value of the target component. Different labels correspond to different parts, so that the machine can conveniently identify and judge.
S400, performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information.
In one embodiment, the generated label graph is morphologically processed, elements of the target component are extracted, information of the outline, the relative position, the overlapping degree and the scale of the target component is obtained, and the information is determined as the information to be compared of the component. Similarly, the same operation and processing are carried out on the standard label graph, the morphological processing is carried out on the component standard label graph, the elements of the target component are extracted, the outline, the relative position, the overlapping degree and the scale information of the target component are obtained, and the information is determined as the information to be compared of the component. After morphological processing is carried out, a part of noise and interference can be removed, and better image information can be obtained.
S500, comparing the information to be compared with the standard comparison information of the component to obtain a comparison result, and judging whether the comparison result is within a set error range.
In one embodiment, the comparison result is obtained by comparing the component to-be-compared information with the component standard comparison information, because the vehicle standard image is a standard non-modified appearance of the vehicle appearance in the vehicle image as described in step S100, the component standard comparison information is used as a comparison reference, because the photograph is different from the actual object, such as a pixel, a color, and the like, a comparison error can be set, and the comparison result is compared with the set error to determine whether the comparison result is within the set error range.
S600, analyzing the comparison result, and judging whether the vehicle appearance is modified or not.
Obtaining a comparison result in the step S500, and if the comparison result is within a set error range, determining that the vehicle appearance is not modified; if the comparison result is not within the set error range, the vehicle appearance is considered to have refitting, the vehicle appearance does not accord with the picture appearance reserved in the vehicle inspection, and the vehicle inspection is unqualified.
The vehicle appearance modification judgment method provided by the embodiment of the application can be faster and more concise to judge whether modification exists on the vehicle appearance, so that the heavy workload of manual examination and verification is reduced, and the working efficiency is properly improved.
In an embodiment, referring to fig. 2, S400 in the method for determining refitting of vehicle appearance further includes the following steps:
s402, setting a comparison reference object, wherein the reference object is a mark post, and acquiring the actual physical size of the mark post.
In one embodiment, in the vehicle appearance judgment, there is a post, which is a reference object placed beside the vehicle when the vehicle picture is taken. The actual size of the target is fixed, so that the actual physical size of the target can be obtained as a reference size.
S404, calculating the proportion of the height of the marker post in the marker post label graph to the actual physical size to obtain a first proportion, and calculating the proportion of the height of the marker post in the marker post standard label to the actual physical size to obtain a second proportion.
In one embodiment, the marker post can also be used as a reference object, and a marker post label graph of the marker post is extracted. The height of the label is calculated in the marker post label graph, and the height can be represented by a pixel value. And calculating the ratio of the height of the marker post in the marker post label graph to the actual physical size to obtain a first ratio, wherein the first ratio is used for calculating the actual size of the target component in the vehicle image. And similarly, calculating the ratio of the height of the marker post in the marker post standard label to the actual physical size to obtain a second ratio, wherein the second ratio is used for calculating the actual size of the target component in the vehicle standard image.
S406, acquiring a first size of the target component in the component label diagram and a second size of the target component in the component standard label diagram.
In one embodiment, a first dimension of the target part in the part tag map may be calculated based on a pixel size of the target part in the part tag map, the first dimension being used to calculate an actual dimension of the target part in the vehicle image. Similarly, a second dimension of the target component in the component standard label map is calculated according to the pixel size of the target component in the component standard label map, and the second dimension is used for calculating the actual dimension of the target component in the vehicle image.
S408, calculating the information to be compared of the component according to the first size and the first proportion, and calculating the standard comparison information of the component according to the second size and the second proportion.
In one embodiment, the actual size of the target component in the vehicle image and the vehicle standard image is calculated based on the first scale and the second scale obtained in step S404, and the first size and the second size obtained in step S406. And calculating a first actual size of the target component according to the first size and the first proportion, and taking the first actual size as the information to be compared. And similarly, calculating a second actual size of the target component according to the second size and the second proportion, and taking the second actual size as standard comparison information.
In an embodiment, referring to fig. 3, S400 in the vehicle exterior refitting determination method may include the steps of:
s420, performing morphological processing on the target area of the part label map, removing noise points, and combining the same targets in the part label map to obtain the target part interesting area label map only containing the target part interesting area.
In one embodiment, the area of the target component in the component label may be the target area, as the label map may have noise, interference, etc. Therefore, morphological processing is performed on the label map, noise points are removed, the same targets in the component label map are merged, and the target component region-of-interest label map only including the target component region-of-interest is obtained in the component label map.
S440, performing morphological processing on the component standard label map, removing noise points, and combining the same targets in the component standard label map to obtain a target component interested area standard label map only containing the target component interested area.
In one embodiment, the area of the target component in the component standard label may be the target area due to factors such as noise that may be present in the label map. Therefore, morphological processing is performed on the standard label map, noise points are removed, and the same targets in the component standard label map are combined to obtain the target component interested area standard label map only containing the target component interested area.
In an embodiment, referring to fig. 4, S420 in the vehicle exterior refitting determination method may include the steps of:
s422, extracting a first region of interest from the target component region of interest label map, and calculating the first size of the target component according to the first region of interest.
In an embodiment, referring to fig. 5, S440 of the method for determining refitting of vehicle appearance may include the steps of:
s444, extracting a second region of interest from the standard label map of the region of interest of the target component, and calculating the second size of the target component according to the second region of interest.
In an embodiment, in the method for judging vehicle appearance modification, step S400 performs morphological processing on the component tag map, extracts the contour, relative position, overlapping degree, and scale information of the target component, obtains component comparison-waiting information, performs morphological processing on the component standard tag map, extracts the contour, relative position, overlapping degree, and scale information of the target component, and obtains component standard comparison information, including;
and S410, the target component is a tail sign board.
In one embodiment, the target component may be selected to be a tailgate panel for the determination of whether a refitting of the vehicle appearance exists.
S412, extracting a tail sign board area from the tail sign board label map, performing binarization and morphological processing on the tail sign board area, counting the number of color changes and the color level of pixel values in the area, and obtaining tail sign board comparison information.
In one embodiment, the area of the vehicle tail marking plate is extracted from the vehicle tail marking plate label image, and binarization processing is performed on the area of the vehicle tail marking plate, so that the image identification is simpler and more convenient. The vehicle tail marking plate is a stripe with alternate colors, so that the external characteristics of the vehicle tail marking plate can be determined by counting the change times of the colors in the region and the color gradation of the pixel values, and the characteristics are used as the comparison information of the vehicle tail marking plate.
And S414, extracting the area of the vehicle tail sign board from the vehicle tail sign board standard label image, carrying out binarization and morphological processing on the vehicle tail sign board area, and counting the number of color changes and the color level of pixel values in the area to obtain the vehicle tail sign board standard information.
And extracting the area of the vehicle tail sign plate in the vehicle tail sign plate standard label image, and performing binarization processing on the area of the vehicle tail sign plate, wherein binarization can make the image identification simpler and more convenient. The vehicle tail marking plate is a stripe with alternate colors, so that the external characteristics of the vehicle tail marking plate can be determined by counting the change times of the colors in the region and the color gradation of the pixel values, and the characteristics are used as the standard comparison information of the vehicle tail marking plate.
In an embodiment, referring to fig. 6, step S600 in the method for determining refitting of vehicle appearance analyzes the comparison result to determine whether there is refitting of the vehicle appearance, including:
s602, the target component is at least one target component, each target component corresponds to one comparison result, and if a plurality of target components are provided, a plurality of comparison results are obtained.
In one embodiment, when determining the appearance of the vehicle, the required target components may be selected according to the actual requirements of the user, and one or more target components may be selected for comparison.
S604, if any comparison result does not meet the preset condition, judging that the vehicle is modified, and judging that the final result is false.
In one embodiment, if a plurality of target components are selected as the determination criteria, a plurality of comparison results are obtained, and if any one of the comparison results does not meet the preset condition, the vehicle appearance is determined to have refitting. The preset condition may be a set error, or data such as a comparison threshold, which may be set according to an actual requirement, and is not limited herein.
And S606, if all comparison results meet the preset conditions, judging that the vehicle is not modified, and obtaining a true final result.
In one embodiment, if a plurality of target components are selected as the determination criteria, a plurality of comparison results are obtained, and if all of the comparison results satisfy the predetermined condition, it is determined that there is no refit in the vehicle appearance. The preset condition may be a set error, or data such as a comparison threshold, which may be set according to an actual requirement, and is not limited herein.
In an embodiment, a semantic segmentation model can be adopted to segment the vehicle image to obtain the vehicle mask map, and the standard image can be segmented to obtain the vehicle standard mask map;
in particular, the semantic segmentation model is adopted to segment the image, the segmentation precision is high, and the segmentation precision can be accurate to pixels.
Referring to FIG. 7, the semantic segmentation model training process is as follows:
s702, a vehicle image training set is obtained.
The vehicle images with different angles, different illumination, different types and different image qualities need to be acquired so as to acquire more data samples, ensure that the vehicle is in the central area of the image as far as possible, and keep the integrity, so that the training model can be more complete. And the data diversity is increased, and the robustness of the model is improved.
And S704, marking the vehicle outline in the vehicle image, setting the values of all pixel points outside the vehicle outline to be 0, and extracting the vehicle area image.
In the vehicle image, marking the outer contour of the vehicle, setting all background pixel points outside the outer contour of the vehicle to be 0, and simultaneously extracting the vehicle in the vehicle image area to obtain a vehicle area image.
S706, marking the outline of the 22 types of target parts in the vehicle area image, setting corresponding label values (1-22) for all pixel points in the outline of the 22 types of target parts, and marking to obtain a target part pixel label image. The vehicle appearance component is not only of 22 types, and according to the needs of actual conditions, 22 types of appearance components needing training can be selected, the types and the number of the components are not limited, here, 22 types are only an embodiment, and the 22 types of appearance components are as follows: [1] the vehicle comprises a vehicle body, a [2] side surface, a [3] lower ventilation area, a [4] upper ventilation area, a [5] front main lamp, a [6] engine hood, a [7] wheel, a [8] side window, a [9] front and rear window, a [10] license plate, a [11] rear main lamp, a [12] fence, a [13] fence plate, a [14] right side protection device, a [15] rear lower protection device, a [16] reflective mark-red, a [17] reflective mark-white, a [18] carriage spray-painted license plate, a [19] tail mark plate, a [20] rail, a [21] reflective mark-yellow and a [22] conical mark rod. And setting label values 1-22 for 22 types of appearance parts respectively to obtain a marked target part pixel label map.
And S708, graying the labeled target component pixel label image, forcibly modifying the value of the edge pixel point by using a nearest neighbor algorithm, generating a label pair with the original image of the vehicle image, and putting the label pair into a semantic segmentation model for training.
The graying process can reduce the original data amount of the image, and is convenient for the subsequent processing with less calculation amount, so that the graying process needs to be carried out on the colorful vehicle image. And forcibly modifying the value of the contour edge pixel point of the target component by using a nearest neighbor algorithm.
S710, preferentially learning foreground and background areas of the whole vehicle in the training process, then finely adjusting the super-parameters and the class weights of the network, and finally obtaining a semantic segmentation model.
In the training process, the network model preferentially learns the foreground and background areas of the whole vehicle in the image, and by the method, the convergence of the model can be accelerated, so that the training process is faster. Then, the weighting of the hyper-parameters and the category of the network is adjusted, and the processes are continuously repeated until the hyper-parameters and the category of the network reach a balance state.
It should be understood that, although the steps in the above-described flowcharts are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in the above-described flowcharts may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the sub-steps or the stages is not necessarily sequential, but may be performed alternately or alternatingly with other steps or at least a portion of the sub-steps or stages of other steps.
In another embodiment, the present invention further provides a vehicle exterior refitting determination device 800, referring to fig. 8, including:
the acquisition module 810, the acquisition module 810 acquires a vehicle image and a vehicle standard image corresponding to the vehicle image;
the segmentation module 820 is used for segmenting the vehicle image to obtain a vehicle mask map, and segmenting the standard image to obtain a vehicle standard mask map;
a label map generating module 830, wherein the label map generating module 830 extracts a component sub-mask map of the target component from the vehicle mask map, generates a component label map, extracts a component standard sub-mask map of the target component from the vehicle standard mask map, and generates a component standard label map;
the image processing module 840 performs morphological processing on the part tag map, extracts the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performs morphological processing on the part standard tag map, extracts the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information;
a comparison module 850, wherein the comparison module 850 compares the component to-be-compared information with the component standard comparison information to obtain a comparison result, and determines whether the comparison result is within a set error range;
and a judging module 860, wherein the judging module 860 analyzes the comparison result and judges whether the vehicle appearance is modified.
In an embodiment, referring to fig. 9, the image processing module 840 in the vehicle exterior refitting determination device 800 includes the following modules:
the calculating module 841, the calculating module 841 calculates the ratio of the height of the post in the post label graph to the actual physical dimension to obtain a first ratio, and calculates the ratio of the height of the post in the post standard label to the actual physical dimension to obtain a second ratio;
a size obtaining module 842, wherein the size obtaining module 842 obtains a first size of the target component on the component label map and a second size of the target component on the component standard label map;
the calculating module 841 calculates the information to be compared of the component according to the first size and the first ratio, and calculates the standard comparison information of the component according to the second size and the second ratio.
In an embodiment, referring to fig. 9, the image processing module 840 in the vehicle exterior refitting determination device 800 includes the following modules:
a morphology processing module 843, where the morphology processing module 843 performs morphology processing on the target area of the part label map, removes noise points, and merges the same targets in the part label map to obtain the target part region-of-interest label map only including the target part region-of-interest; and performing morphological processing on the component standard label map, removing noise points, and combining the same targets in the component standard label map to obtain the target component interested area standard label map only containing the target component interested area.
In one embodiment, the morphology processing module 843 in the vehicle appearance modification determination device 800 includes the following modules:
an extraction module that extracts a first region of interest in the target component region of interest label map; and extracting a second region of interest from the standard label map of the region of interest of the target component.
The calculation module 841 calculates the first size of the target component based on a first region of interest and calculates the second size of the target component based on a second region of interest.
In one embodiment, the target component is a tail sign; in the tag map of the vehicle tail sign board, an extraction module extracts the area of the vehicle tail sign board, a morphology processing module 843 performs binarization and morphology processing on the area of the vehicle tail sign board, and an image processing module 840 counts the number of times of color change in the area and the color level of a pixel value to obtain comparison information of the vehicle tail sign board; in the standard tag map of the vehicle tail marking plate, an extraction module extracts the area of the vehicle tail marking plate, a morphology processing module 843 performs binarization and morphology processing on the area of the vehicle tail marking plate, and an image processing module 840 counts the number of times of color change and the color level of a pixel value in the area to obtain standard comparison information of the vehicle tail marking plate; the comparison module 850 compares the vehicle tail sign board comparison information with the vehicle tail sign board standard information, and the judgment module judges whether the vehicle tail sign board is qualified.
In an embodiment, the determining module 860 analyzes the comparison result, and determining whether there is a modification in the vehicle appearance further includes: the target component is at least one target component, each target component corresponds to one comparison result, and if the target components are multiple, multiple comparison results are obtained; if any comparison result does not meet the preset condition, judging that the vehicle appearance is modified, and judging that the final result is false; and if all the comparison results meet the preset conditions, judging that the appearance of the vehicle is not modified, and obtaining a true final result.
In an embodiment, referring to fig. 10, the segmentation module 820 in the vehicle appearance modification determination apparatus 800 includes a generation module 821 for obtaining a vehicle image training set; marking a vehicle outline in the vehicle image, setting the values of all pixel points outside the vehicle outline to be 0, and extracting a vehicle area image; marking the outline of a 22-type target component in the vehicle area image, setting corresponding label values (1-22) for all pixel points in the outline of the 22-type target component, and obtaining a target component pixel label image after marking; graying the labeled image of the target component pixel, forcibly modifying the value of the edge pixel point by using a nearest neighbor algorithm, generating a label pair with the original image of the vehicle image, and putting the label pair into a semantic segmentation model for training; in the training process, the foreground and background areas of the whole vehicle need to be preferentially learned, then the super-parameters and the category weights of the network are finely adjusted, and finally the semantic segmentation model is obtained.
For specific limitations of the vehicle appearance detection device, reference may be made to the above limitations of the vehicle appearance detection method, which are not described herein again. The respective modules in the vehicle appearance detecting apparatus described above may be implemented in whole or in part by software, hardware, and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
The invention also provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method when executing the computer program. The computer device may be an end product, the processor of which is used to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a vehicle refitting determination method.
The invention also provides a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method.
Those skilled in the art will appreciate that all or part of the processes of the methods of the above embodiments may be implemented by a computer program that instructs associated hardware to perform the processes of the embodiments of the methods described above. Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, Phase-Change RAM (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), flash Memory or other Memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, magnetic cassette tape, tape-Disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A vehicle refitting determination method, the method comprising:
acquiring a vehicle image and a vehicle standard image corresponding to the vehicle image;
segmenting the vehicle image to obtain a vehicle mask image, and segmenting the standard image to obtain a vehicle standard mask image;
extracting a component sub-mask diagram of a target component from the vehicle mask diagram to generate a component label diagram, and extracting a component standard sub-mask diagram of the target component from the vehicle standard mask diagram to generate a component standard label diagram;
performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information;
comparing the information to be compared with the standard comparison information of the component to obtain a comparison result, and judging whether the comparison result is within a set error range;
and analyzing the comparison result, and judging whether the vehicle appearance is modified or not.
2. The vehicle exterior refitting judgment method according to claim 1, wherein the morphological processing of the part label map to extract the contour, relative position, overlapping degree, and dimension information of the target part to obtain part comparison-waiting information, the morphological processing of the standard label map to extract the contour, relative position, overlapping degree, and dimension information of the target part to obtain part comparison-standard information comprises:
setting a comparison reference object, wherein the reference object is a mark post, and acquiring the actual physical size of the mark post;
calculating the ratio of the height of the marker post in the marker post label graph to the actual physical size to obtain a first ratio, and calculating the ratio of the height of the marker post in the marker post standard label to the actual physical size to obtain a second ratio;
acquiring a first size of the target component in the component label diagram and a second size of the target component in the component standard label diagram;
and calculating the information to be compared of the component according to the first size and the first proportion, and calculating the standard comparison information of the component according to the second size and the second proportion.
3. The vehicle exterior refitting judgment method according to claim 2, wherein the morphological processing of the part label map to extract the contour, relative position, overlapping degree and scale information of the target part to obtain part comparison-waiting information, the morphological processing of the standard label map to extract the contour, relative position, overlapping degree and scale information of the target part to obtain part comparison-standard information comprises:
performing morphological processing on the target area of the part label map, removing noise points, and combining the same targets in the part label map to obtain a target part interesting area label map only containing the target part interesting area;
and performing morphological processing on the component standard label map, removing noise points, and combining the same targets in the component standard label map to obtain the target component interested area standard label map only containing the target component interested area.
4. The vehicle exterior refitting determination method according to claim 3, wherein the method further comprises:
extracting a first region of interest from the target component region of interest label map, and calculating the first size of the target component according to the first region of interest;
and extracting a second region of interest from the standard label map of the region of interest of the target component, and calculating the second size of the target component according to the second region of interest.
5. The vehicle exterior refitting judgment method according to claim 1, wherein the morphological processing of the part label map to extract the contour, relative position, overlapping degree, and dimension information of the target part to obtain part comparison-waiting information, the morphological processing of the standard label map to extract the contour, relative position, overlapping degree, and dimension information of the target part to obtain part comparison-standard information comprises:
the target component is a vehicle tail sign board;
extracting a tail sign board area from the tail sign board label map, carrying out binarization and morphological processing on the tail sign board area, counting the number of times of color change in the area and the color level of a pixel value, and obtaining tail sign board comparison information;
and extracting a vehicle tail sign board area from the vehicle tail sign board standard label map, performing binarization and morphological processing on the vehicle tail sign board area, counting the number of times of color change in the area and the color level of a pixel value, and obtaining vehicle tail sign board standard comparison information.
6. The vehicle exterior refitting determination method according to claim 1,
analyzing the comparison result, and judging whether the vehicle appearance is modified or not comprises the following steps:
the target component is at least one target component, each target component corresponds to one comparison result, and if the target components are multiple, multiple comparison results are obtained;
if any comparison result does not meet the preset condition, judging that the vehicle appearance is modified, and judging that the final result is false;
and if all the comparison results meet the preset conditions, judging that the appearance of the vehicle is not modified, and obtaining a true final result.
7. The vehicle exterior refitting determination method according to claim 1,
adopting a semantic segmentation model to segment the vehicle image to obtain the vehicle mask image, and segmenting the standard image to obtain the vehicle standard mask image;
the semantic segmentation model training process is as follows:
acquiring a vehicle image training set;
marking a vehicle outline in the vehicle image, setting the values of all pixel points outside the vehicle outline to be 0, and extracting a vehicle area image;
marking the outline of the 22 types of target components in the vehicle area image, setting corresponding label values for all pixel points in the outline of the 22 types of target components, and obtaining a target component pixel label image after marking;
graying the labeled image of the target component pixel, forcibly modifying the value of the edge pixel point by using a nearest neighbor algorithm, generating a label pair with the original image of the vehicle image, and putting the label pair into a semantic segmentation model for training;
in the training process, the foreground and background areas of the whole vehicle need to be preferentially learned, then the super-parameters and the category weights of the network are finely adjusted, and finally the semantic segmentation model is obtained.
8. A vehicle exterior refitting determination device, comprising:
the acquisition module acquires a vehicle image and a vehicle standard image corresponding to the vehicle image;
the segmentation module is used for segmenting the vehicle image to obtain a vehicle mask image and segmenting the standard image to obtain a vehicle standard mask image;
the tag map generation module extracts a component sub-mask map of a target component from the vehicle mask map to generate a component tag map, extracts a component standard sub-mask map of the target component from the vehicle standard mask map to generate a component standard tag map;
the image processing module is used for performing morphological processing on the part label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part comparison information, performing morphological processing on the part standard label graph, extracting the outline, the relative position, the overlapping degree and the scale information of the target part to obtain part standard comparison information;
the comparison module compares the information to be compared of the component with the component standard comparison information to obtain a comparison result, and judges whether the comparison result is within a set error range;
and the judging module analyzes the comparison result and judges whether the vehicle appearance is modified or not.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911354664.7A CN111126393A (en) | 2019-12-22 | 2019-12-22 | Vehicle appearance refitting judgment method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911354664.7A CN111126393A (en) | 2019-12-22 | 2019-12-22 | Vehicle appearance refitting judgment method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111126393A true CN111126393A (en) | 2020-05-08 |
Family
ID=70503298
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911354664.7A Pending CN111126393A (en) | 2019-12-22 | 2019-12-22 | Vehicle appearance refitting judgment method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111126393A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111461093A (en) * | 2020-06-22 | 2020-07-28 | 北京慧智数据科技有限公司 | Modified truck identification method based on deep learning technology |
| CN111784662A (en) * | 2020-06-29 | 2020-10-16 | 深圳至峰精密制造有限公司 | Workpiece recognition method, workpiece recognition device, computer equipment and storage medium |
| CN112697066A (en) * | 2020-12-02 | 2021-04-23 | 王刚 | Vehicle part positioning method and device and computer storage medium |
| CN113505781A (en) * | 2021-06-01 | 2021-10-15 | 北京旷视科技有限公司 | Target detection method and device, electronic equipment and readable storage medium |
| CN114140754A (en) * | 2021-11-30 | 2022-03-04 | 北京比特易湃信息技术有限公司 | Refitted vehicle identification method based on deep learning |
| CN114743120A (en) * | 2022-06-10 | 2022-07-12 | 深圳联和智慧科技有限公司 | Roadside vehicle illegal lane occupation detection method and system based on image recognition |
| WO2023123818A1 (en) * | 2021-12-30 | 2023-07-06 | 上海商汤智能科技有限公司 | Vehicle retrofitting detection method and apparatus, electronic device, computer-readable storage medium and computer program product |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110047771A1 (en) * | 2009-09-03 | 2011-03-03 | Cabot Smith | Method of converting a hybrid sedan into a hybrid livery vehicle |
| WO2018028230A1 (en) * | 2016-08-10 | 2018-02-15 | 东方网力科技股份有限公司 | Deep learning-based method and device for segmenting vehicle license plate characters, and storage medium |
| CN110245691A (en) * | 2019-05-27 | 2019-09-17 | 上海眼控科技股份有限公司 | A kind of intelligent identification Method of vehicle appearance color discoloration repacking |
| CN110378236A (en) * | 2019-06-20 | 2019-10-25 | 西安电子科技大学 | Testing vehicle register identification model construction, recognition methods and system based on deep learning |
-
2019
- 2019-12-22 CN CN201911354664.7A patent/CN111126393A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110047771A1 (en) * | 2009-09-03 | 2011-03-03 | Cabot Smith | Method of converting a hybrid sedan into a hybrid livery vehicle |
| WO2018028230A1 (en) * | 2016-08-10 | 2018-02-15 | 东方网力科技股份有限公司 | Deep learning-based method and device for segmenting vehicle license plate characters, and storage medium |
| CN110245691A (en) * | 2019-05-27 | 2019-09-17 | 上海眼控科技股份有限公司 | A kind of intelligent identification Method of vehicle appearance color discoloration repacking |
| CN110378236A (en) * | 2019-06-20 | 2019-10-25 | 西安电子科技大学 | Testing vehicle register identification model construction, recognition methods and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 闫世宇;何岭松;尤若;高志强;罗瀛;: "基于手机平台的工件台残余切屑检测方法" * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111461093A (en) * | 2020-06-22 | 2020-07-28 | 北京慧智数据科技有限公司 | Modified truck identification method based on deep learning technology |
| CN111784662A (en) * | 2020-06-29 | 2020-10-16 | 深圳至峰精密制造有限公司 | Workpiece recognition method, workpiece recognition device, computer equipment and storage medium |
| CN112697066A (en) * | 2020-12-02 | 2021-04-23 | 王刚 | Vehicle part positioning method and device and computer storage medium |
| CN113505781A (en) * | 2021-06-01 | 2021-10-15 | 北京旷视科技有限公司 | Target detection method and device, electronic equipment and readable storage medium |
| CN114140754A (en) * | 2021-11-30 | 2022-03-04 | 北京比特易湃信息技术有限公司 | Refitted vehicle identification method based on deep learning |
| WO2023123818A1 (en) * | 2021-12-30 | 2023-07-06 | 上海商汤智能科技有限公司 | Vehicle retrofitting detection method and apparatus, electronic device, computer-readable storage medium and computer program product |
| CN114743120A (en) * | 2022-06-10 | 2022-07-12 | 深圳联和智慧科技有限公司 | Roadside vehicle illegal lane occupation detection method and system based on image recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111126393A (en) | Vehicle appearance refitting judgment method and device, computer equipment and storage medium | |
| CN113160192B (en) | Visual sense-based snow pressing vehicle appearance defect detection method and device under complex background | |
| CN106875381B (en) | Mobile phone shell defect detection method based on deep learning | |
| CN108898047B (en) | Pedestrian detection method and system based on blocking and shielding perception | |
| US11080839B2 (en) | System and method for training a damage identification model | |
| CN104778721B (en) | The distance measurement method of conspicuousness target in a kind of binocular image | |
| CN111325769B (en) | Target object detection method and device | |
| CN105574550A (en) | Vehicle identification method and device | |
| CN105447512A (en) | Coarse-fine optical surface defect detection method and coarse-fine optical surface defect detection device | |
| CN110569856A (en) | sample labeling method and device, and damage category identification method and device | |
| CN106557740B (en) | The recognition methods of oil depot target in a kind of remote sensing images | |
| CN114926441B (en) | Defect detection method and system for injection molding part machining and forming | |
| CN104615986A (en) | Method for utilizing multiple detectors to conduct pedestrian detection on video images of scene change | |
| CN110288612B (en) | Nameplate positioning and correcting method and device | |
| Galsgaard et al. | Circular hough transform and local circularity measure for weight estimation of a graph-cut based wood stack measurement | |
| CN108764338B (en) | Pedestrian tracking method applied to video analysis | |
| CN111784675A (en) | Method and device for processing article texture information, storage medium and electronic equipment | |
| WO2020047316A1 (en) | System and method for training a damage identification model | |
| CN108664970A (en) | A kind of fast target detection method, electronic equipment, storage medium and system | |
| CN112784712B (en) | Missing child early warning implementation method and device based on real-time monitoring | |
| Wang et al. | Combining semantic scene priors and haze removal for single image depth estimation | |
| CN111402185B (en) | Image detection method and device | |
| CN112581495A (en) | Image processing method, device, equipment and storage medium | |
| CN111507177B (en) | Identification method and device for metering turnover cabinet | |
| CN114373144B (en) | Automatic identification method for circular identification points in high-speed video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20240913 |