CN106778778A - A kind of high-speed hardware multiple target feature extracting method - Google Patents
A kind of high-speed hardware multiple target feature extracting method Download PDFInfo
- Publication number
- CN106778778A CN106778778A CN201611091297.2A CN201611091297A CN106778778A CN 106778778 A CN106778778 A CN 106778778A CN 201611091297 A CN201611091297 A CN 201611091297A CN 106778778 A CN106778778 A CN 106778778A
- Authority
- CN
- China
- Prior art keywords
- module
- multiple target
- data
- target feature
- shape detection
- 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
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
- G06V10/40—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
- G06V10/443—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 by matching or filtering
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J13/00—Controls for manipulators
- B25J13/08—Controls for manipulators by means of sensing devices, e.g. viewing or touching devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of high-speed hardware multiple target feature extracting method, including photographing module, SHAPE DETECTION module, image characteristics extraction module, hardware identification sort module and handling module, the high-speed hardware multiple target feature extracting method is comprised the following steps；S1, photographing module is arranged on the opening end of material production line, and the shooting end of photographing module is aligned into conveyer belt；Be filmed for article on production line by S2, the photographing module in the S1, and is connected by data wire, and view data is passed into SHAPE DETECTION module, and SHAPE DETECTION module is analyzed to data.In the present invention, FPGA algorithms are applied to when on production line, for prior art, the technology can realize that the multiple target shape of hardware-level is quickly recognized, FPGA algorithms are realized for computer realization, with low cost, low in energy consumption, stable grade is a little；Production line efficiency is greatly improved, a large amount of artificial and times are saved.
Description
Technical field
The present invention relates to target's feature-extraction method and technology field, more particularly to a kind of high-speed hardware multiple target feature extraction
Method.
Background technology
Production line is exactly the route that process of producing product is passed through, i.e., since raw material into production scene, by processing,
The routes that a series of production production line activities such as transport, assembling, inspection are constituted.Production line is organized by the principle of object
, complete a kind of productive organization of Product Process process, i.e., by product specialization principle, outfit turns out a produce（0th,
Part）Required various equipment and the workman of each work post, are responsible for completing certain product（Parts and components）Whole manufacture work,
The processing of different process is carried out to the identical subject of labour, now with the development of society and science and technology, increasing article is all
Start, using automatic production line next life product product, to greatly reduce the use of manpower, and also improve the efficiency of production, portion
Divide and often judge to be operated according to the shape facility of article sometimes in automatic production line, and automatic production line is automatic at present
SHAPE DETECTION technical efficiency is low, is typically only capable to operate in that tens of frames are per second, and comparatively production efficiency be not very high, be this I
Propose a kind of high-speed hardware multiple target feature extracting method.
The content of the invention
Based on the technical problem that background technology is present, the present invention proposes a kind of high-speed hardware multiple target feature extraction side
Method.
A kind of high-speed hardware multiple target feature extracting method proposed by the present invention, including photographing module, SHAPE DETECTION module,
Image characteristics extraction module, hardware identification sort module and handling module, the high-speed hardware multiple target feature extracting method include
Following steps；
S1, photographing module is arranged on the opening end of material production line, and the shooting end of photographing module is aligned into conveyer belt, and
And photographing module is shot using the pixel of N × N；
Be filmed for article on production line by S2, the photographing module in the S1, and is connected by data wire, by picture number
According to SHAPE DETECTION module is passed to, SHAPE DETECTION module is analyzed to data, and the SHAPE DETECTION module will shoot what is obtained
Image is divided into M × M cell；
Data after analysis are passed to image characteristics extraction module by S3, the SHAPE DETECTION module in the S2 by data wire,
The image moment computing module that described image characteristic extracting module is contained within is calculated the characteristics of image of each cell, then
Operated instead of pixel execution flag using the image moment of unit；
S4, image characteristics extraction module is by extracting zeroth order square, first moment and the high-order autocorrelation haracter of each object in the S3
Levy, and be marked, then the data of extraction are passed to hardware identification classification mould by image characteristics extraction module by data wire
Data after last hardware identification sort module splits data into L region, and are made unit mark figure by block；
S5, the data transfer that the hardware identification sort module in the S4 will have been analyzed gives long-range user's control end, user's control
Data are preserved and pass to handling module by end processed, and handling module is positioned according to the data for obtaining to target item, and is carried out
Fast and effectively capture, so repeatedly, you can complete the extraction to high-speed hardware multiple target feature.
Preferably, in the S4, while execution flag is operated, in addition it is also necessary to which the image for updating each target at any time is special
Levy.
Preferably, the photographing module, SHAPE DETECTION module, image characteristics extraction module, hardware identification sort module and
Contain the High Speed General algorithm routine gone out using FPRG technological development in handling module.
Preferably, the photographing module and SHAPE DETECTION module are installed in the opening end of production line, and handling module is pacified
At the downstream of production line.
Preferably, the handling module is manipulator.
Preferably, in the S5, manipulator crawl scope is reached according to location information and streamline prediction of speed material
Time, send and be accurately controlled signal, control machinery hand carries out grasping manipulation to material.
In the present invention, FPGA employs logical cell array LCA（Logic Cell Array）Such a concept, it is internal
Including configurable logic blocks CLB（Configurable Logic Block）, input/output module IOB（Input Output
Block）And interconnector（Interconnect）Three parts.Field programmable gate array（FPGA）It is programming device, with
Conventional logic circuit and gate array（Such as PAL, GAL and CPLD devices）Compare, FPGA has different structures.FPGA is using small-sized
Look-up table（16×1RAM）To realize combinational logic, each look-up table is connected to an input for d type flip flop, and trigger comes again
Drive other logic circuits or drive I/O, to thus constitute and can not only realize combination logic function but also can realize sequential logic function
Basic logic unit module, these intermodules are interconnected using metal connecting line or are connected to I/O modules.The logic of FPGA is
Programming data is loaded by internally static storage cell to realize, storage value in a memory cell determines logic list
Between the logic function and each module of unit or the connecting mode between module and I/O, and finally determine achieved by FPGA
Function, FPGA allows unlimited number of programming, and FPGA is to set its working condition by the program being stored in ram in slice, because
This, needs to be programmed the RAM in piece during work.User can be according to different configuration modes, using different programming sides
Formula, during power-up, fpga chip reads in piece in programming RAM data in EPROM, and after the completion of configuration, FPGA enters working condition.
After power down, FPGA reverts to white tiles, and internal logic relation disappears, therefore, FPGA being capable of Reusability.The programming of FPGA need not be specially
FPGA programmable devices, need only use general EPROM, PROM programmable device.When modification FPGA functions are needed, one only need to be changed
Piece EPROM.So, with a piece of FPGA, different programming datas can produce different circuit functions.Therefore, FPGA
Using very flexibly, and this technology is applied to when on production line, for prior art, the technology can realize hardware
The multiple target shape of rank quickly recognizes that FPGA algorithms are realized for computer realization, low in energy consumption with low cost, fortune
Row stabilization etc. is a little；Production line efficiency is greatly improved, a large amount of artificial and times are saved.
Specific embodiment
The present invention is made with reference to specific embodiment further explain.
Reference picture 1, a kind of high-speed hardware multiple target feature extracting method proposed by the present invention, including photographing module, shape
Detection module, image characteristics extraction module, hardware identification sort module and handling module, and, photographing module, SHAPE DETECTION mould
Contain the height gone out using FPRG technological development in block, image characteristics extraction module, hardware identification sort module and handling module
Fast general-purpose algorithm program, the high-speed hardware multiple target feature extracting method is comprised the following steps；
S1, photographing module is arranged on the opening end of material production line, and the shooting end of photographing module is aligned into conveyer belt, and
And photographing module is shot using the pixel of N × N；
Be filmed for article on production line by S2, the photographing module in the S1, and is connected by data wire, by picture number
According to SHAPE DETECTION module is passed to, SHAPE DETECTION module is analyzed to data, and the SHAPE DETECTION module will shoot what is obtained
Image is divided into M × M cell；
Data after analysis are passed to image characteristics extraction module by S3, the SHAPE DETECTION module in the S2 by data wire,
The image moment computing module that described image characteristic extracting module is contained within is calculated the characteristics of image of each cell, then
Operated instead of pixel execution flag using the image moment of unit；
S4, image characteristics extraction module is by extracting zeroth order square, first moment and the high-order autocorrelation haracter of each object in the S3
Levy, and be marked, in addition, while execution flag is operated, in addition it is also necessary to update the characteristics of image of each target at any time, then
The data of extraction are passed to hardware identification sort module, last hardware identification classification by image characteristics extraction module by data wire
After module splits data into L region, and data are made unit mark figure；
S5, the data transfer that the hardware identification sort module in the S4 will have been analyzed gives long-range user's control end, user's control
Data are preserved and pass to handling module by end processed, and handling module is manipulator, and photographing module and SHAPE DETECTION module are equal
Installed in the opening end of production line, handling module is arranged at the downstream of production line, and handling module is according to the data for obtaining to mesh
Mark article positioning, the time that manipulator captures scope is reached according to location information and streamline prediction of speed material, sends accurate
Control signal, control machinery hand carries out grasping manipulation to material, so repeatedly, you can complete to high-speed hardware multiple target feature
Extraction.
The above, the only present invention preferably specific embodiment, but protection scope of the present invention is not limited thereto,
Any one skilled in the art the invention discloses technical scope in, technology according to the present invention scheme and its
Inventive concept is subject to equivalent or change, should all be included within the scope of the present invention.
Claims (6)
1. a kind of high-speed hardware multiple target feature extracting method, including photographing module, SHAPE DETECTION module, image characteristics extraction mould
Block, hardware identification sort module and handling module, it is characterised in that the high-speed hardware multiple target feature extracting method includes following
Step；
S1, photographing module is arranged on the opening end of material production line, and the shooting end of photographing module is aligned into conveyer belt, and
And photographing module is shot using the pixel of N × N；
Be filmed for article on production line by S2, the photographing module in the S1, and is connected by data wire, by picture number
According to SHAPE DETECTION module is passed to, SHAPE DETECTION module is analyzed to data, and the SHAPE DETECTION module will shoot what is obtained
Image is divided into M × M cell；
Data after analysis are passed to image characteristics extraction module by S3, the SHAPE DETECTION module in the S2 by data wire,
The image moment computing module that described image characteristic extracting module is contained within is calculated the characteristics of image of each cell, then
Operated instead of pixel execution flag using the image moment of unit；
S4, image characteristics extraction module is by extracting zeroth order square, first moment and the high-order autocorrelation haracter of each object in the S3
Levy, and be marked, then the data of extraction are passed to hardware identification classification mould by image characteristics extraction module by data wire
Data after last hardware identification sort module splits data into L region, and are made unit mark figure by block；
S5, the data transfer that the hardware identification sort module in the S4 will have been analyzed gives long-range user's control end, user's control
Data are preserved and pass to handling module by end processed, and handling module is positioned according to the data for obtaining to target item, and is carried out
Fast and effectively capture, so repeatedly, you can complete the extraction to high-speed hardware multiple target feature.
2. a kind of high-speed hardware multiple target feature extracting method according to claim 1, it is characterised in that in the S4,
While execution flag is operated, in addition it is also necessary to update the characteristics of image of each target at any time.
3. a kind of high-speed hardware multiple target feature extracting method according to claim 1, it is characterised in that the shooting mould
Containing using FPRG skills in block, SHAPE DETECTION module, image characteristics extraction module, hardware identification sort module and handling module
The High Speed General algorithm routine that art is developed.
4. a kind of high-speed hardware multiple target feature extracting method according to claim 1, it is characterised in that the shooting mould
Block and SHAPE DETECTION module are installed in the opening end of production line, and handling module is arranged at the downstream of production line.
5. a kind of high-speed hardware multiple target feature extracting method according to claim 1, it is characterised in that the crawl mould
Block is manipulator.
6. a kind of high-speed hardware multiple target feature extracting method according to claim 1, it is characterised in that in the S5,
The time that manipulator captures scope is reached according to location information and streamline prediction of speed material, is sent and is accurately controlled signal,
Control machinery hand carries out grasping manipulation to material.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611091297.2A CN106778778A (en) | 2016-12-01 | 2016-12-01 | A kind of high-speed hardware multiple target feature extracting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611091297.2A CN106778778A (en) | 2016-12-01 | 2016-12-01 | A kind of high-speed hardware multiple target feature extracting method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106778778A true CN106778778A (en) | 2017-05-31 |
Family
ID=58913641
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611091297.2A Pending CN106778778A (en) | 2016-12-01 | 2016-12-01 | A kind of high-speed hardware multiple target feature extracting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106778778A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107291475A (en) * | 2017-06-27 | 2017-10-24 | 中国电子产品可靠性与环境试验研究所 | Universal PHM application configurations method and apparatus |
| CN107590051A (en) * | 2017-09-06 | 2018-01-16 | 郑州云海信息技术有限公司 | A kind of device and method of display RACK server hard disc running statuses |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101872423A (en) * | 2010-05-27 | 2010-10-27 | 天津大学 | Method for tracking moving object on production line |
| CN103617413A (en) * | 2013-11-07 | 2014-03-05 | 电子科技大学 | Method for identifying object in image |
| CN104058260A (en) * | 2013-09-27 | 2014-09-24 | 沈阳工业大学 | Robot automatic stacking method based on visual processing |
| CN104156726A (en) * | 2014-08-19 | 2014-11-19 | 大连理工大学 | Workpiece recognition method based on geometric shape feature and device thereof |
| CN104574408A (en) * | 2015-01-16 | 2015-04-29 | 东华大学 | Industry transparent film package detecting method and device based on shape feature extraction |
-
2016
- 2016-12-01 CN CN201611091297.2A patent/CN106778778A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101872423A (en) * | 2010-05-27 | 2010-10-27 | 天津大学 | Method for tracking moving object on production line |
| CN104058260A (en) * | 2013-09-27 | 2014-09-24 | 沈阳工业大学 | Robot automatic stacking method based on visual processing |
| CN103617413A (en) * | 2013-11-07 | 2014-03-05 | 电子科技大学 | Method for identifying object in image |
| CN104156726A (en) * | 2014-08-19 | 2014-11-19 | 大连理工大学 | Workpiece recognition method based on geometric shape feature and device thereof |
| CN104574408A (en) * | 2015-01-16 | 2015-04-29 | 东华大学 | Industry transparent film package detecting method and device based on shape feature extraction |
Non-Patent Citations (2)
| Title |
|---|
| 牛文生: "《机载计算机技术》", 31 January 2013 * |
| 苏阳等: "《那些年，我们拿下了FPGA》", 31 October 2013 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107291475A (en) * | 2017-06-27 | 2017-10-24 | 中国电子产品可靠性与环境试验研究所 | Universal PHM application configurations method and apparatus |
| CN107291475B (en) * | 2017-06-27 | 2020-06-12 | 中国电子产品可靠性与环境试验研究所 | Universal PHM application configuration method and device |
| CN107590051A (en) * | 2017-09-06 | 2018-01-16 | 郑州云海信息技术有限公司 | A kind of device and method of display RACK server hard disc running statuses |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12131524B2 (en) | Systems and methods for optical material characterization of waste materials using machine learning | |
| US9796120B2 (en) | Injection molding system | |
| CN109483573A (en) | Machine learning device, robot system and machine learning method | |
| CN106778778A (en) | A kind of high-speed hardware multiple target feature extracting method | |
| CN106062819B (en) | Image processing apparatus, imaging sensor, image processing method | |
| CN113902901B (en) | Object separation method and system based on lightweight detection | |
| Shill et al. | Plant disease detection based on YOLOv3 and YOLOv4 | |
| CN111008561A (en) | Livestock quantity determination method, terminal and computer storage medium | |
| US20240189866A1 (en) | System and method for sorting objects | |
| Babu et al. | Design and development of cost effective arduino based object sorting system | |
| CN203803804U (en) | Rapid identification card sorting instrument | |
| CN110175506A (en) | Pedestrian based on parallel dimensionality reduction convolutional neural networks recognition methods and device again | |
| KR102490912B1 (en) | Server and controlling method thereof | |
| Vo et al. | Automating Tomato Ripeness Classification and Counting with YOLOv9 | |
| JP2007200309A (en) | Position specification method and apparatus for plant target part, and working robot using the same apparatus | |
| Vijayakanthan et al. | Classification of vegetable plant pests using deep transfer learning | |
| CN106920255B (en) | Moving object extraction method and device for image sequence | |
| CN112488135A (en) | Method for classifying BIM three-dimensional image features of transformer substation | |
| CN117773913A (en) | Robot control method and device, storage medium and robot | |
| CN105934721A (en) | Data management device and data management program | |
| CN111353577A (en) | Optimization method and device of multi-task-based cascade combination model and terminal equipment | |
| CN209787572U (en) | Automatic online plug-in control system and plug-in mechanism | |
| US10731984B2 (en) | Reconfigurable sensor unit for electronic device | |
| Anandh et al. | IOT based Automated Object Colour Sorting and Counting System | |
| Debnárová et al. | Group technology in context of the product classification |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170531 |