CN113572997A - Video stream data analysis method, device, equipment and storage medium - Google Patents
Video stream data analysis method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN113572997A CN113572997A CN202110829141.4A CN202110829141A CN113572997A CN 113572997 A CN113572997 A CN 113572997A CN 202110829141 A CN202110829141 A CN 202110829141A CN 113572997 A CN113572997 A CN 113572997A
- Authority
- CN
- China
- Prior art keywords
- video stream
- video frame
- stream data
- video
- analysis
- 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
- 238000007405 data analysis Methods 0.000 title claims abstract description 110
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000004458 analytical method Methods 0.000 claims abstract description 126
- 238000004590 computer program Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 12
- 238000012545 processing Methods 0.000 description 9
- 238000012544 monitoring process Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 3
- 239000000779 smoke Substances 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003139 buffering effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/76—Television signal recording
- H04N5/91—Television signal processing therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computational Linguistics (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The embodiment of the invention discloses a video stream data analysis method, a video stream data analysis device, video stream data analysis equipment and a storage medium. The method comprises the following steps: acquiring multi-channel video stream data in real time; storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool; and acquiring a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images respectively. The technical scheme realizes real-time intelligent analysis on large-scale multi-channel video stream data.
Description
Technical Field
The embodiment of the invention relates to the technical field of image processing, in particular to a video stream data analysis method, a video stream data analysis device, video stream data analysis equipment and a storage medium.
Background
In modern life, with the enhancement of economic development and security awareness, video monitoring is undoubtedly an indispensable ring in security and protection. In addition, with the development of machine learning technology, the monitoring video gradually realizes intellectualization, and meanwhile, the video analysis also enters large scale.
At present, most of video intelligent identification aims at single-path or a few paths of videos, and analysis results are stored by using persistence processing after video frames are intelligently analyzed. However, how to implement real-time intelligent analysis of large-scale multi-channel video streams is an urgent problem to be solved.
Disclosure of Invention
The embodiment of the invention provides a video stream data analysis method, a video stream data analysis device, video stream data analysis equipment and a storage medium, and aims to realize real-time intelligent analysis on large-scale multi-channel video streams.
In a first aspect, an embodiment of the present invention provides a method for analyzing video stream data, including:
acquiring multi-channel video stream data in real time;
storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool;
and acquiring a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images respectively.
Optionally, obtaining a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images, respectively, includes: and inputting a plurality of video frame images acquired in batch in the image buffer pool into a target data analysis model, and obtaining analysis results respectively corresponding to the plurality of video frame images through inference of the target data analysis model.
In the embodiment, a plurality of video frame images are reasoned simultaneously, hardware resources are fully utilized, the inference speed of the deep learning model is accelerated, and therefore the real-time performance of intelligent analysis of large-scale multi-channel video stream data is improved.
Optionally, before acquiring multiple video stream data in real time, the method further includes: and responding to a data analysis model selection request, and taking a data analysis model matched with the data analysis model selection request as the target data analysis model.
In the technical scheme, a plurality of data analysis models are preset, and the target data analysis model can be determined based on the data analysis model selection request so as to be suitable for different application scenes, so that the applicability of the video stream data analysis method is improved.
Optionally, storing the video frame image corresponding to the multiple paths of video stream data into an image buffer pool, including: decoding each path of video stream data respectively to obtain each video frame image corresponding to each path of video stream data; and storing each video frame image corresponding to each path of video stream data into the image buffer pool through the data input queue corresponding to each path of video stream data.
In this embodiment, a data input queue may be respectively set for each path of video stream data, and the decoding result of the video stream data is stored in the image buffer pool through the corresponding data input queue, so that the video data of different paths can be processed separately, and the accuracy of each path of video data is ensured.
Optionally, outputting analysis results respectively corresponding to the plurality of video frame images includes: and outputting analysis results respectively corresponding to the plurality of video frame images through data output queues respectively corresponding to each path of video stream data.
In this embodiment, a data output queue may be respectively set for each path of video stream data, and the corresponding intelligent analysis result of the video stream data is output through the corresponding data output queue, so that the video data of different paths can be processed separately, and the accuracy of each path of video data is ensured.
Optionally, the video frame image and the analysis result corresponding to the video frame image both carry video stream number information and timestamp information.
In the technical scheme, the video frame image and the analysis result thereof carry the video stream number information and the event stamp information, so that the accuracy of each path of video data is ensured, and the problems of confusion among multiple paths of videos and disorder of a single path of video can be avoided.
Optionally, after outputting the analysis results corresponding to the plurality of video frame images, the method further includes: and coding the analysis results respectively corresponding to the plurality of video frame images, and outputting the analysis result video streams respectively corresponding to the plurality of paths of video stream data.
In the technical scheme, the video stream data of the analysis result corresponding to the large-scale multi-channel video stream is output and displayed, so that the large-scale multi-channel video is conveniently monitored in real time.
In a second aspect, an embodiment of the present invention further provides a video stream data analysis apparatus, where the apparatus includes:
the video stream acquisition module is used for acquiring multi-channel video stream data in real time;
the video frame image storage module is used for storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool;
and the intelligent analysis module is used for acquiring a plurality of video frame images in batch in the image buffer pool for analysis and outputting analysis results respectively corresponding to the plurality of video frame images.
In a third aspect, an embodiment of the present invention further provides a computer device, where the computer device includes:
one or more processors;
a memory for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method according to any one of the embodiments of the invention.
In a fourth aspect, the embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the method according to any embodiment of the present invention.
According to the technical scheme of the embodiment of the invention, the multi-channel video stream data acquired in real time is decoded to obtain the video frame images corresponding to the multi-channel video stream data, the video frame images are stored in the image buffer pool, then the plurality of video frame images are acquired in batch in the image buffer pool to be analyzed, and the corresponding analysis results are output, so that the real-time intelligent analysis of the large-scale multi-channel video stream data can be realized. The image buffer pool is favorable for subsequent batch processing of video frame images, and the efficiency of real-time intelligent analysis of large-scale multi-channel video stream data is improved.
Drawings
Fig. 1 is a flowchart of a video stream data analysis method according to a first embodiment of the present invention;
fig. 2 is a flowchart of a video stream data analysis method according to a second embodiment of the present invention;
fig. 3 is a schematic diagram of a system architecture to which a video stream data analysis method according to a second embodiment of the present invention is applied;
fig. 4 is a schematic structural diagram of a video stream data analysis apparatus according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a computer device in the fourth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It is to be further noted that, for the convenience of description, only a part of the structure relating to the present invention is shown in the drawings, not the whole structure.
Example one
Fig. 1 is a flowchart of a video stream data analysis method according to an embodiment of the present invention, where the present embodiment is applicable to a case of performing real-time intelligent analysis on a large-scale multi-channel video stream data, and the method may be performed by a video stream data analysis apparatus, which may be implemented in a hardware and/or software manner, and may be generally integrated in a computer device.
As shown in fig. 1, the method for analyzing video stream data provided in this embodiment includes the following steps:
and S110, acquiring multi-channel video stream data in real time.
Video stream data refers to the transmission of video data, i.e. video data can be processed as a steady continuous stream through the network. The multi-channel video stream data refers to a transport stream of video data collected by a plurality of camera devices (such as cameras).
The type of the video stream data may be RTMP (Real Time Messaging Protocol), or may be other types of video stream formats. Taking the Haikang camera as an example, the monitored content can be automatically converted into the RTSP video stream.
Optionally, the multiple paths of video stream data may be a transmission stream of multiple paths of video data acquired by an intelligent monitoring system, for example, the multiple paths of video stream data may be an intelligent monitoring system applied to traffic, energy, parks, schools, buildings, banks, and small and medium-sized enterprises. Specifically, the number of the paths of the multi-path video stream data may be a relatively large number, for example, the number of the paths is greater than a preset threshold, that is, the multi-path video stream data may be a transport stream of the multi-path video data acquired by a large-scale intelligent monitoring system, and the number of the paths of the video stream data is not specifically limited in this embodiment.
As an optional implementation, when multiple paths of video stream data are acquired, the data flow rate of each path of video stream data in unit time may be monitored, and if the data flow rate of any path of video stream data in unit time is greater than the flow rate threshold, the definition of the path of video stream data may be controlled, for example, the definition of the path of video stream data may be reduced according to a preset amplitude.
And S120, storing the video frame image corresponding to the multi-path video stream data into an image buffer pool.
The image buffer pool refers to a buffer area for buffering images.
And respectively decoding each path of video stream data to obtain a video frame image corresponding to each path of video stream data, and storing the video frame images corresponding to each path of video stream data into an image buffer pool.
As an optional implementation manner, the image buffer pool may be partitioned, and the video frame images corresponding to the video stream data of different paths are stored in different image buffer pool partitions, so as to distinguish the number of video stream paths to which the video frame images stored in the image buffer pool belong.
As another optional implementation manner, before storing the video frame images corresponding to each path of video stream data in the image buffer pool, respectively adding the identity identifiers corresponding to each path of video stream data to the video frame images corresponding to each path of video stream data, so as to distinguish the number of video streams to which the video frame images stored in the image buffer pool belong.
Further, when the video frame images are stored in the image buffer pool, each video frame image may be preprocessed in the image buffer pool, for example, the preprocessing may be image denoising processing, image enhancement processing, image size resetting processing, and the like.
S130, obtaining a plurality of video frame images in the image buffer pool in batch for analysis, and outputting analysis results corresponding to the video frame images respectively.
The method includes obtaining a plurality of video frame images in batch in the image buffer pool, where the plurality of video frame images may be obtained in batch according to an order in which the video frame images are stored in the image buffer pool, or may also be obtained in batch in different partitions of the image buffer pool, which is not limited in this embodiment.
In this embodiment, a plurality of video frame images obtained in batch in the image buffer pool are intelligently analyzed at the same time, so as to obtain intelligent analysis results respectively corresponding to the plurality of video frame images. The intelligent analysis may be, for example, image recognition, object detection, etc. of the video frame image. The image recognition refers to recognizing places, identifications, people, objects, buildings, other variables and the like in an image; the object detection means that the position of an object in an image is marked by using a mark frame, and the class of the object is given.
Illustratively, the method for intelligently analyzing the images of the video frames acquired in batch in the image buffer pool is applicable to retrieval and image enhancement of mass videos by users, and is applicable to capture and recognition of vehicle license plates by intelligent transportation users, especially capture and recognition of vehicle license plates in abnormal weather.
Optionally, the method includes obtaining a plurality of video frame images in batch in the image buffer pool and performing intelligent analysis simultaneously based on a deep learning technology.
As an optional implementation manner, obtaining a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results respectively corresponding to the plurality of video frame images may specifically be:
inputting a plurality of video frame images acquired in batch in the image buffer pool into a target data analysis model, and obtaining analysis results respectively corresponding to the plurality of video frame images through inference of the target data analysis model.
The target data analysis model refers to any depth learning model for intelligently analyzing video frame images. For example, the target data analysis model may be a face detection analysis model for performing face recognition in the video frame image; the target data analysis model can also be an object detection model used for detecting objects in the video frame image; the target data analysis model can also be a face structural attribute model, and is used for carrying out face structural attribute identification in the video frame image, such as head portrait hair style identification, whether glasses are worn or not and the like; the target data analysis model can also be a vehicle intelligent recognition model and is used for vehicle recognition, vehicle license plate recognition and the like in the video frame image; the target data analysis model can also be a flame smoke intelligent detection model and is used for carrying out smoke and fire detection and the like in a video frame image; the target data analysis model can also be a human body attribute intelligent identification model used for identifying human body attributes in video frame images, such as identifying human body gender, human body clothes and the like.
The target data analysis model may be a deep learning network model supporting parameter adjustment, and the size of the target data analysis model is not specifically limited in this embodiment.
And inputting a plurality of video frame images obtained in batch in the image buffer pool into a target data analysis model, performing image reasoning through the target data analysis model, and outputting to obtain analysis results corresponding to the plurality of video frame images.
Optionally, an application program interface corresponding to the target data analysis model is called, a data analysis request is sent to the target data analysis model, and a data analysis response fed back by the application program interface is received. Wherein the data analysis request comprises a plurality of video frame images acquired in the image buffer pool; the data analysis response comprises intelligent analysis results of the target data analysis model aiming at the video frame images.
In this embodiment, the target data analysis model performs intelligent analysis on a plurality of video frame images acquired in batch in the image buffer pool at the same time. The target data analysis model may be a model inference of a neural network using a GPU (Graphics Processing Unit) Graphics card.
In the embodiment, the GPU is used for reasoning a plurality of video frame images simultaneously, so that hardware resources are fully utilized, the reasoning speed of the deep learning model is accelerated, and the real-time performance of the intelligent analysis of the large-scale multi-channel video stream data is improved.
As an optional implementation manner, before acquiring multiple paths of video stream data in real time, the method may further include:
and responding to a data analysis model selection request, and taking a data analysis model matched with the data analysis model selection request as the target data analysis model.
The data analysis model selection request refers to a request for selecting one data analysis model from a plurality of data analysis models as a target data analysis model. The data analysis model selection request may be initiated by a user before the video stream data analysis method provided in this embodiment is executed according to an actual application requirement. The data analysis model selection request may be, for example, a request for using a face detection analysis model as the target data analysis model, so that the video stream data analysis method provided in this embodiment may implement face recognition on large-scale multi-channel video stream data.
In this embodiment, a plurality of data analysis models may be preset for performing intelligent analysis on video stream data, for example, a face detection analysis model, an object detection model, a face structured attribute model, a vehicle intelligent recognition model, a flame and smoke intelligent detection model, a human body attribute intelligent recognition model, and the like may be preset.
Wherein, one calling application program interface can be independently set for each data analysis model. The calling of the target data analysis model can be realized through a calling application program interface corresponding to the target data analysis model, and further intelligent analysis matched with the target data analysis model is performed on large-scale multi-channel video stream data.
As an optional implementation manner, an application program interface may be provided for presetting a data analysis model to implement enriching video intelligent analysis types supported by the video stream data analysis method provided by this embodiment according to actual application requirements. The preset data analysis model is carried out through the application program interface, and can be directly called, so that corresponding intelligent analysis on large-scale multi-channel video stream data is realized.
After the data analysis model selection request is received, analyzing parameters carried by the data analysis model selection request, and selecting a matched data analysis model from a plurality of preset data analysis models as the target data analysis model.
In the technical scheme, a plurality of data analysis models are preset, and the target data analysis model can be determined based on the data analysis model selection request so as to be suitable for different application scenes, so that the applicability of the video stream data analysis method is improved.
As an optional implementation manner, after outputting the analysis results corresponding to the plurality of video frame images, the method may further include: and coding the analysis results respectively corresponding to the plurality of video frame images, and outputting the analysis result video streams respectively corresponding to the plurality of paths of video stream data.
The analysis result video stream refers to video stream data carrying intelligent analysis results. Taking the target data analysis model as the object detection model as an example, the analysis result video stream may be video stream data identified with the object detection frame.
After obtaining the analysis results of the plurality of video frame images corresponding to each path of video stream data, the image analysis results corresponding to each path of video stream data may be encoded and the analysis result video streams corresponding to each path of video stream data may be output.
When the analysis results of a plurality of video frame images corresponding to any path of video stream data are coded, the coding can be carried out according to the data type of the corresponding path of video stream, so that the data of the analysis result video stream is consistent with the data type of the corresponding video stream data. The data type may be webRTC (Web Real-Time Communication, from Web instant messaging), HLS (HTTP Live Streaming, HTTP-based adaptive bitrate Streaming protocol), RTMP-FLV (RTMP-FLASH VIDEO), and the like.
Optionally, the analysis result video stream corresponding to each path of video stream data may be output to the front end, and the analysis result video streams corresponding to each path of video stream data are decoded and displayed uniformly by the front end, so as to monitor a large-scale multi-path video. The display of the analysis result video stream can be realized in a WEB page mode.
In the technical scheme, the video stream data of the analysis result corresponding to the large-scale multi-channel video stream is output and displayed, so that the large-scale multi-channel video is conveniently monitored in real time.
According to the technical scheme of the embodiment of the invention, the multi-channel video stream data acquired in real time is decoded to obtain the video frame images corresponding to the multi-channel video stream data, the video frame images are stored in the image buffer pool, then the plurality of video frame images are acquired in batch in the image buffer pool to be analyzed, and the corresponding analysis results are output, so that the real-time intelligent analysis of the large-scale multi-channel video stream data can be realized. The image buffer pool is favorable for subsequent batch processing of video frame images, and the efficiency of real-time intelligent analysis of large-scale multi-channel video stream data is improved.
Example two
Fig. 2 is a flowchart of a video stream data analysis method according to a second embodiment of the present invention, which is embodied on the basis of the foregoing embodiment, where the storing of the video frame images corresponding to the multiple paths of video stream data in the image buffer pool may specifically be:
decoding each path of video stream data respectively to obtain each video frame image corresponding to each path of video stream data; and storing each video frame image corresponding to each path of video stream data into the image buffer pool through the data input queue corresponding to each path of video stream data.
Wherein, outputting the analysis results respectively corresponding to the plurality of video frame images may specifically be: and outputting analysis results respectively corresponding to the plurality of video frame images through data output queues respectively corresponding to each path of video stream data.
As shown in fig. 2, the method for analyzing video stream data provided in this embodiment includes the following steps:
and S210, acquiring multi-channel video stream data in real time.
And S220, respectively decoding each path of video stream data to obtain each video frame image corresponding to each path of video stream data.
And S230, storing each video frame image corresponding to each path of video stream data into the image buffer pool through the data input queue corresponding to each path of video stream data.
In this embodiment, a data input queue is generated for each video for storing and transmitting image data. For each path of video stream data, after video stream decoding is carried out, the obtained video frame images are sent into corresponding data input queues frame by frame and sent into an image buffer pool through the data input queues. The data input queue may be a first-in first-out queue.
In this embodiment, a data input queue may be respectively set for each path of video stream data, and the decoding result of the video stream data is stored in the image buffer pool through the corresponding data input queue, so that the video data of different paths can be processed separately, and the accuracy of each path of video data is ensured.
As an optional implementation manner, the video frame image obtained by decoding the video stream carries video stream number information and timestamp information.
The video stream number information is used for indicating the number of paths to which the video stream belongs; time stamp information for indicating that it is time information of the video frame image in the video stream data.
The video stream number information and the data input queue number information are in one-to-one correspondence, and the video stream number information and the data output queue number information are also in one-to-one correspondence. Alternatively, the video stream number information, the data input queue number information, and the data output queue number information corresponding to the same video stream data may be the same.
Correspondingly, the analysis result corresponding to the video frame image also carries video stream number information and time stamp information.
The video stream number information and the time stamp information of the analysis result corresponding to any one video frame image are the same as those of the video frame image.
Before and after each video frame image is analyzed, a video stream number information is given to each video frame image, and the video stream number information is used for determining a data input queue and a data output queue corresponding to each video frame image. That is, each video frame image, in its category, will include two parameters to indicate its source. Illustratively, the image category of a certain video frame image includes the following parameters: { "camera Id": cam-01"," timestamp ":1488627991133}, where camera Id is the video stream number (ID) and timestamp is the timestamp of video frame generation. Accordingly, the image category of the analysis result of the video frame image still includes the following parameters: { "camera Id": cam-01"," timestamp ":1488627991133 }.
In the technical scheme, the video frame image and the analysis result thereof carry the video stream number information and the event stamp information, so that the accuracy of each path of video data is ensured, and the problems of confusion among multiple paths of videos and disorder of a single path of video can be avoided.
S240, obtaining a plurality of video frame images in batch in the image buffer pool for analysis.
The image buffer pool can reserve a large number of video frame images, analyze and arrange the images in the video frame images, and fill the images into a target data analysis model for model reasoning.
The plurality of video frame images obtained in batch in the image buffer pool can be input into a target data analysis model, and analysis results respectively corresponding to the plurality of video frame images are obtained through inference by the target data analysis model.
And S250, outputting analysis results corresponding to the video frame images through data output queues corresponding to the video stream data of each path.
In this embodiment, a data output queue is generated for each video for storing and transmitting image data.
And the analysis result corresponding to each video frame image still carries the video stream number information and the timestamp information, and is sent to a corresponding data output queue according to the video stream number information and is output through the corresponding data output queue. The data output queue may be a first-in first-out queue.
In this embodiment, a data output queue may be respectively set for each path of video stream data, and the corresponding intelligent analysis result of the video stream data is output through the corresponding data output queue, so that the video data of different paths can be processed separately, and the accuracy of each path of video data is ensured.
And S260, coding analysis results corresponding to the plurality of video frame images respectively, and outputting analysis result video streams corresponding to the plurality of paths of video stream data respectively.
And coding the analysis result corresponding to the video frame image output in each data output queue to respectively obtain the analysis result data video corresponding to each path of video stream data. And transmitting each path of analysis result data video to the front end, and decoding and displaying.
Fig. 3 shows a system architecture to which the video stream data analysis method provided in this embodiment is applied, and as shown in fig. 3, the system architecture includes n data input queues, an image buffer pool, an intelligent analysis module, and n data output queues. Referring to fig. 3, multiple paths of video stream data are acquired in real time, after each path of video data is decoded, a corresponding video frame image is input into a corresponding data input queue, and the video frame image is sent into an image buffer pool through the data input queue, so that a large number of video frame images are stored in the image buffer pool. The intelligent analysis model obtains a plurality of video frame images in batch in the image buffer pool, calls the target data analysis model to perform intelligent analysis, inputs analysis results corresponding to the video frame images into corresponding data output queues, outputs the analysis results through the data output queues, transmits video streams corresponding to the analysis results after encoding the analysis results, and then decodes and displays the video streams through the front end, so that real-time monitoring of the video analysis results can be realized.
For the parts of this embodiment that are not explained in detail, reference may be made to the aforementioned embodiments, which are not described herein again.
By the aid of the technical scheme, real-time intelligent analysis and identification of large-scale multi-channel videos are achieved, and the large-scale multi-channel videos are converted into the analysis result video streams to be output in real time, and if the analysis result video streams can be displayed in a web front-end page, real-time monitoring of video analysis results is facilitated, and monitoring of intelligent analysis results by a user is more convenient.
EXAMPLE III
Fig. 4 is a schematic structural diagram of a video stream data analysis apparatus according to a third embodiment of the present invention, where the third embodiment is applicable to a case of performing real-time intelligent analysis on large-scale multiple paths of video stream data, and the apparatus may be implemented in a software and/or hardware manner, and may be generally integrated in a computer device. As shown in fig. 4, the video stream data analysis apparatus specifically includes: a video stream acquisition module 310, a video frame image storage module 320, and an intelligent analysis module 330. Wherein,
a video stream acquiring module 310, configured to acquire multiple paths of video stream data in real time;
a video frame image storage module 320, configured to store video frame images corresponding to the multiple paths of video stream data into an image buffer pool;
and the intelligent analysis module 330 is configured to obtain a plurality of video frame images in batch in the image buffer pool for analysis, and output analysis results corresponding to the plurality of video frame images respectively.
According to the technical scheme of the embodiment of the invention, the multi-channel video stream data acquired in real time is decoded to obtain the video frame images corresponding to the multi-channel video stream data, the video frame images are stored in the image buffer pool, then the plurality of video frame images are acquired in batch in the image buffer pool to be analyzed, and the corresponding analysis results are output, so that the real-time intelligent analysis of the large-scale multi-channel video stream data can be realized.
Optionally, the intelligent analysis module 330 is specifically configured to input a plurality of video frame images obtained in batch in the image buffer pool into a target data analysis model, and obtain analysis results corresponding to the plurality of video frame images through inference by the target data analysis model.
Further, the above apparatus further comprises: and the data analysis model selection module is used for responding to a data analysis model selection request before acquiring the multi-channel video stream data in real time, and taking a data analysis model matched with the data analysis model selection request as the target data analysis model.
Optionally, the video frame image storage module 320 is specifically configured to decode each path of video stream data respectively to obtain each video frame image corresponding to each path of video stream data; and storing each video frame image corresponding to each path of video stream data into the image buffer pool through the data input queue corresponding to each path of video stream data.
Optionally, the video frame image storage module 320 is specifically configured to output analysis results corresponding to the plurality of video frame images through data output queues corresponding to each path of video stream data respectively.
Optionally, the video frame image and the analysis result corresponding to the video frame image both carry video stream number information and timestamp information.
Optionally, the apparatus further comprises: and the analysis result output module is used for encoding the analysis results respectively corresponding to the plurality of video frame images after outputting the analysis results respectively corresponding to the plurality of video frame images, and outputting the analysis result video streams respectively corresponding to the plurality of paths of video stream data.
The video stream data analysis device provided by the embodiment of the invention can execute the video stream data analysis method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
Example four
Fig. 5 is a schematic structural diagram of a computer apparatus according to a fourth embodiment of the present invention, as shown in fig. 5, the computer apparatus includes a processor 410, a memory 420, an input device 430, and an output device 440; the number of processors 410 in the computer device may be one or more, and one processor 410 is taken as an example in fig. 5; the processor 410, the memory 420, the input device 430 and the output device 440 in the computer apparatus may be connected by a bus or other means, and the connection by the bus is exemplified in fig. 5.
The memory 420 serves as a computer-readable storage medium, and may be used to store software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the video stream data analysis method in the embodiment of the present invention (for example, the video stream acquisition module 310, the video frame image storage module 320, and the intelligent analysis module 330 in the video stream data analysis apparatus). The processor 410 executes various functional applications and data processing of the computer device by executing software programs, instructions and modules stored in the memory 420, that is, implements the video stream data analysis method described above.
The memory 420 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 420 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 420 may further include memory located remotely from processor 410, which may be connected to a computer device through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 430 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the computer apparatus. The output device 440 may include a display device such as a display screen.
EXAMPLE five
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a method for analyzing video stream data, the method including:
acquiring multi-channel video stream data in real time;
storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool;
and acquiring a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images respectively.
Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the method operations described above, and may also perform related operations in the video stream data analysis method provided by any embodiment of the present invention.
From the above description of the embodiments, it is obvious for a person skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the video stream data analysis apparatus, the units and modules included in the embodiment are merely divided according to the functional logic, but are not limited to the above division as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions without departing from the scope of the invention. Therefore, although the present invention has been described in more detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (10)
1. A method for analyzing video stream data, comprising:
acquiring multi-channel video stream data in real time;
storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool;
and acquiring a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images respectively.
2. The method according to claim 1, wherein obtaining a plurality of video frame images in batch in the image buffer pool for analysis, and outputting analysis results corresponding to the plurality of video frame images respectively comprises:
inputting a plurality of video frame images acquired in batch in the image buffer pool into a target data analysis model, and obtaining analysis results respectively corresponding to the plurality of video frame images through inference of the target data analysis model.
3. The method of claim 2, further comprising, prior to acquiring the plurality of video stream data in real-time:
and responding to a data analysis model selection request, and taking a data analysis model matched with the data analysis model selection request as the target data analysis model.
4. The method of claim 1, wherein storing the video frame images corresponding to the plurality of video stream data into an image buffer pool comprises:
decoding each path of video stream data respectively to obtain each video frame image corresponding to each path of video stream data;
and storing each video frame image corresponding to each path of video stream data into the image buffer pool through the data input queue corresponding to each path of video stream data.
5. The method according to claim 1 or 4, wherein outputting the analysis results corresponding to the plurality of video frame images respectively comprises:
and outputting analysis results respectively corresponding to the plurality of video frame images through data output queues respectively corresponding to each path of video stream data.
6. The method of claim 5, wherein the video frame image and the analysis result corresponding to the video frame image each carry video stream number information and timestamp information.
7. The method according to claim 1, further comprising, after outputting the analysis results corresponding to the plurality of video frame images, respectively:
and coding the analysis results respectively corresponding to the plurality of video frame images, and outputting the analysis result video streams respectively corresponding to the plurality of paths of video stream data.
8. A video stream data analysis apparatus, comprising:
the video stream acquisition module is used for acquiring multi-channel video stream data in real time;
the video frame image storage module is used for storing the video frame images corresponding to the multi-channel video stream data into an image buffer pool;
and the intelligent analysis module is used for acquiring a plurality of video frame images in batch in the image buffer pool for analysis and outputting analysis results respectively corresponding to the plurality of video frame images.
9. A computer device, characterized in that the computer device comprises:
one or more processors;
a memory for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110829141.4A CN113572997A (en) | 2021-07-22 | 2021-07-22 | Video stream data analysis method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110829141.4A CN113572997A (en) | 2021-07-22 | 2021-07-22 | Video stream data analysis method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113572997A true CN113572997A (en) | 2021-10-29 |
Family
ID=78166247
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110829141.4A Pending CN113572997A (en) | 2021-07-22 | 2021-07-22 | Video stream data analysis method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113572997A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866436A (en) * | 2014-06-12 | 2015-08-26 | 国家电网公司 | Method for storing massive security incidents |
| CN109711323A (en) * | 2018-12-25 | 2019-05-03 | 武汉烽火众智数字技术有限责任公司 | A kind of live video stream analysis accelerated method, device and equipment |
| CN109769115A (en) * | 2019-01-04 | 2019-05-17 | 武汉烽火众智数字技术有限责任公司 | A kind of method, apparatus and equipment of Intelligent Optimal video analysis performance |
| CN110365942A (en) * | 2019-07-18 | 2019-10-22 | 北京工业大数据创新中心有限公司 | A kind of real-time video intelligent analysis method and system |
| CN110765288A (en) * | 2019-09-04 | 2020-02-07 | 北京旷视科技有限公司 | Image information synchronization method, device and system and storage medium |
| CN111046074A (en) * | 2019-12-13 | 2020-04-21 | 北京百度网讯科技有限公司 | Streaming data processing method, device, equipment and medium |
| CN111723727A (en) * | 2020-06-17 | 2020-09-29 | 平安科技（深圳）有限公司 | Cloud monitoring method and device based on edge computing, electronic equipment and storage medium |
| CN112989078A (en) * | 2021-03-26 | 2021-06-18 | 北京市商汤科技开发有限公司 | Data processing method and device, computer equipment and storage medium |
-
2021
- 2021-07-22 CN CN202110829141.4A patent/CN113572997A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866436A (en) * | 2014-06-12 | 2015-08-26 | 国家电网公司 | Method for storing massive security incidents |
| CN109711323A (en) * | 2018-12-25 | 2019-05-03 | 武汉烽火众智数字技术有限责任公司 | A kind of live video stream analysis accelerated method, device and equipment |
| CN109769115A (en) * | 2019-01-04 | 2019-05-17 | 武汉烽火众智数字技术有限责任公司 | A kind of method, apparatus and equipment of Intelligent Optimal video analysis performance |
| CN110365942A (en) * | 2019-07-18 | 2019-10-22 | 北京工业大数据创新中心有限公司 | A kind of real-time video intelligent analysis method and system |
| CN110765288A (en) * | 2019-09-04 | 2020-02-07 | 北京旷视科技有限公司 | Image information synchronization method, device and system and storage medium |
| CN111046074A (en) * | 2019-12-13 | 2020-04-21 | 北京百度网讯科技有限公司 | Streaming data processing method, device, equipment and medium |
| CN111723727A (en) * | 2020-06-17 | 2020-09-29 | 平安科技（深圳）有限公司 | Cloud monitoring method and device based on edge computing, electronic equipment and storage medium |
| CN112989078A (en) * | 2021-03-26 | 2021-06-18 | 北京市商汤科技开发有限公司 | Data processing method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111723727B (en) | Cloud monitoring method and device based on edge computing, electronic equipment and storage medium | |
| EP2688296B1 (en) | Video monitoring system and method | |
| CN107197384B (en) | The multi-modal exchange method of virtual robot and system applied to net cast platform | |
| WO2022227490A1 (en) | Behavior recognition method and apparatus, device, storage medium, computer program, and program product | |
| WO2022041830A1 (en) | Pedestrian re-identification method and device | |
| CN112965804A (en) | Method, device, terminal, system and storage medium for processing information | |
| WO2016000429A1 (en) | Method and device for detecting video conference hotspot scenario | |
| CN112528898A (en) | Alarm event aggregation method and device based on multi-target detection of surveillance videos | |
| CN109803112A (en) | Video analysis management method based on big data, apparatus and system, storage medium | |
| CN114679607B (en) | Video frame rate control method and device, electronic equipment and storage medium | |
| CN112149551A (en) | Safety helmet identification method based on embedded equipment and deep learning | |
| CN112419639A (en) | Video information acquisition method and device | |
| CN111586432A (en) | Method and device for determining air-broadcast live broadcast room, server and storage medium | |
| CN113572997A (en) | Video stream data analysis method, device, equipment and storage medium | |
| CN113516082A (en) | Detection method and device of safety helmet, computer equipment and storage medium | |
| CN108600254A (en) | A kind of audio and video identifying system | |
| CN113395523A (en) | Image decoding method, device and equipment based on parallel threads and storage medium | |
| CN114220175B (en) | Motion pattern recognition method and device, equipment, medium and product thereof | |
| CN116132623A (en) | Intelligent analysis method, system and equipment based on video monitoring | |
| CN113794601B (en) | Network traffic processing method, device and computer readable storage medium | |
| CN114495188B (en) | Image data processing method and device and related equipment | |
| CN112419638B (en) | Method and device for acquiring alarm video | |
| CN111401317B (en) | Video classification method, device, equipment and storage medium | |
| US10650843B2 (en) | System and method for processing sound beams associated with visual elements | |
| CN113938707A (en) | Video processing method, recording and playing box and computer readable storage medium |
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 |