CN112380954A - Video classification intercepting system and method based on image recognition - Google Patents
Video classification intercepting system and method based on image recognition Download PDFInfo
- Publication number
- CN112380954A CN112380954A CN202011247468.2A CN202011247468A CN112380954A CN 112380954 A CN112380954 A CN 112380954A CN 202011247468 A CN202011247468 A CN 202011247468A CN 112380954 A CN112380954 A CN 112380954A
- Authority
- CN
- China
- Prior art keywords
- video
- image recognition
- video image
- module
- classification
- 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 description 25
- 238000012300 Sequence Analysis Methods 0.000 claims abstract description 13
- 238000004458 analytical method Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims description 39
- 238000001514 detection method Methods 0.000 claims description 33
- 238000000605 extraction Methods 0.000 claims description 13
- 238000012795 verification Methods 0.000 claims description 12
- 238000013145 classification model Methods 0.000 claims description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 2
- 239000012634 fragment Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a video classification intercepting system based on image recognition, which comprises: the video image capturing module is used for capturing a video image according to a certain frequency, zooming the video image into a preset image and detecting the size of resolution; the video image recognition module is used for calling a video image recognition model which is trained in advance, obtaining the content of the appointed single-frame video image through the video image capturing module, and sending the content into the video image recognition model to obtain the video image recognition result; the video image recognition result collection module, the video image recognition sequence analysis module and the video analysis and interception module are used for intercepting a video according to the timestamp corresponding to the recognition category sequence, acquiring video clips of the video and finally acquiring a plurality of video clips corresponding to different categories. The system can effectively solve the problem of the singleness of the traditional video classification, and obtain a plurality of label information of the video classification and a corresponding sub-video content system.
Description
Technical Field
The invention relates to the technical field of video image processing, in particular to a video classification intercepting system and method based on image recognition.
Background
With the development and application of computer technology and artificial intelligence image recognition, a computer can recognize multimedia files such as images and videos by executing a related algorithm and can also classify the multimedia files so as to reduce the workload of manual classification and improve the working efficiency of people. The conventional video classification method is used for identifying the whole video file to finally obtain a single classification label, cannot describe detailed video information, and cannot effectively extract highlight sub-video content in a video. Namely, the traditional method ignores the diversity of videos and cannot effectively identify and extract the highlight sub-videos in the videos. In the existing method, a single classification label is simply obtained through a model, and a time node for intercepting the category wonderful sub-video cannot be obtained through the model.
Therefore, the conventional video classification method has the following problems: the whole video content is directly sent to the recognition model to obtain a category result, the problem of diversity of the video content is ignored, a single label content is simply obtained, and further effective information extraction is not performed on the video content, such as extraction of a highlight sub-video. But for video content, the user is not only interested in the classification information of a single video, but may be more in line with the user interest range for key video clips in the video.
Disclosure of Invention
The invention aims to overcome the defects in the background art and provide a video classification intercepting system and method based on image recognition, wherein a model recognition mode is changed to classify video pictures to obtain a time sequence related to a video classification result so as to obtain a plurality of label contents of the video, the time range of the video of the category is extracted by analyzing the time sequence of the video classification result so as to finally obtain a plurality of classification labels and a plurality of corresponding sub-video contents of the video, the problem of the singleness of the traditional video classification can be effectively solved, and a plurality of label information and corresponding sub-video contents of the video classification can be obtained.
In order to achieve the technical effects, the invention adopts the following technical scheme:
a video classification intercepting system based on image recognition comprises:
the video image capturing module is used for capturing a video image according to a certain frequency, zooming the video image into a preset image and detecting the size of resolution;
the video image recognition module is used for calling a video image recognition model which is trained in advance, obtaining the content of the appointed single-frame video image through the video image capturing module, and sending the content into the video image recognition model to obtain the video image recognition result;
the video image recognition result collection module is used for initializing the length of the sequencer according to the original video length and the detection frequency and storing the video image recognition result according to a time sequence;
the video image identification sequence analysis module is used for completing detection and identification of the whole video according to a preset detection frequency, obtaining a detection result corresponding to each frame of image according to a time sequence, and traversing the sequencer to obtain time nodes corresponding to continuous types and types of corresponding videos;
the video analyzing and intercepting module is used for intercepting a video according to the timestamp corresponding to the identification category sequence, acquiring video clips of the video and finally acquiring a plurality of video clips corresponding to different categories;
the video image capturing module, the video image recognition result collecting module, the video image recognition sequence analyzing module and the video analyzing and intercepting module are sequentially connected;
the video classification intercepting system based on image identification obtains a video image identification result time sequence by identifying an ordered video image content set, analyzes the video image identification result time sequence to obtain video key sub-video fragment node information and the fragment category information, obtains a plurality of key sub-videos and corresponding key sub-video category information according to the node information through a video analysis intercepting module, finally shows the key sub-video content of a complete video, and marks the complete video with multi-type label information.
Further, the video image identification model is a convolutional neural network model.
Further, the video image recognition model adopts a mobile net series network structure of google for a feature extraction network during feature extraction.
Further, the training process of the convolutional neural network model is as follows:
A. collecting image samples of specified resolutions of a plurality of categories, and setting network structure input;
B. generating tfrecrd data format supported by tensorflow aiming at picture sample information and picture samples, and dividing the tfrecrd data format into a training file and a verification file according to a certain proportion, wherein the images of the training file and the verification file are different, and the stored image format is the same as the label information;
C. training the model by using a training file to generate a picture content classification model with appointed N classification contents, and verifying the picture content classification model by using a verification file;
D. and judging whether the model loss is reduced to a preset threshold or the training steps reach the preset steps, finishing the training if the model loss is reduced to the preset threshold or the training steps reach the preset steps, and increasing video content picture samples or adjusting model parameters and returning to the step A if the model loss is not reduced to the preset threshold and the training steps do not reach the preset steps until the training is finished.
Further, the step a further includes performing normalization processing on each collected image sample, which is beneficial to accelerating convergence of the model.
Meanwhile, the invention also discloses a video classification intercepting method based on image recognition, which is realized by the video classification intercepting system based on image recognition and comprises the following steps:
s1, a video image capturing module captures a video image according to a certain frequency, zooms the video image into a preset picture and detects the size of resolution;
s2, calling a video image recognition model trained in advance by a video image recognition module, obtaining the content of a specified single-frame video image through a video image capturing module, and sending the content into the video image recognition model to obtain a video image recognition result;
s3, initializing the length of the sequencer by the video image identification result collection module through the original video length and the detection frequency, and storing the video image identification result according to the time sequence;
s4, a video image identification sequence analysis module completes detection and identification of the whole video according to a preset detection frequency, obtains a detection result corresponding to each frame image according to a time sequence, and a traversal sequencer obtains time nodes corresponding to continuous types and types of corresponding videos;
s5, intercepting the video by a video analyzing and intercepting module according to the timestamp corresponding to the identification category sequence, acquiring video clips of the video, and finally acquiring a plurality of video clips corresponding to different categories;
the video classification intercepting method based on image recognition can recognize various types of video information and extract and intercept the key sub-video content of the corresponding type, utilizes the image recognition technology to recognize and process a video image set, obtains the video type and video type node information by analyzing a recognition result sequence, finally obtains a plurality of key sub-video contents and a plurality of types of labels, can be effectively applied to the fields of video classification and video interception, realizes the automatic labeling of video files and the extraction of video highlights key videos, meets the requirements of the automatic video classification technology and the video intercepting method, and effectively improves the effectiveness and diversity of the extraction of the video content information.
Further, the step S3 is specifically:
the video image recognition result collection module calculates the length of the sequencer through the video time length and the detection frequency, and obtains and stores the result into the image recognition sequencer according to the time sequence by calculating the video image result under the appointed frequency until the video detection is finished or the sequencer is filled.
Further, the step S4 is specifically:
the video image recognition sequence analysis module traverses the image recognition sequencer, stores the current position N when recognizing the video type P, and records the current position M when recognizing the non-video type P, so that the current video clip is the sub-video type P from the time N to the time M;
and finally outputting the starting time and the ending time of a plurality of video segments and marking corresponding class labels according to the mode until the image recognition sequencer is traversed.
Further, the step S5 is specifically:
the video analyzing and intercepting module obtains time nodes corresponding to a plurality of image recognition sequencers, calculates the real time length of a corresponding video according to the video length and the detection frequency, cuts the complete video into a plurality of video segments, and prints a plurality of category labels on the complete video;
and informing the user of the corresponding video category label through video recommendation analysis, and displaying the cut video segment to the user.
Compared with the prior art, the invention has the following beneficial effects:
according to the video classification intercepting system and method based on image recognition, the sub-video clips of different types and various types of videos are obtained by analyzing the video image content set, so that not only can the videos be automatically labeled, but also a user can obtain the key sub-video clip content without completely watching the whole video, the diversity of video file labels is increased, the highlight clip information of the videos is effectively extracted and effectively recommended and introduced, the problem of singleness of the traditional video classification is effectively solved, and a plurality of label information of the video classification and the corresponding sub-video content are obtained.
Drawings
FIG. 1 is a schematic diagram of a video classification intercepting system based on image recognition of the invention.
FIG. 2 is a schematic diagram of video image recognition sequence analysis according to one embodiment of the present invention.
Detailed Description
The invention will be further elucidated and described with reference to the embodiments of the invention described hereinafter.
Example (b):
the first embodiment is as follows:
as shown in fig. 1, a video classification intercepting system based on image recognition includes: the device comprises a video image capturing module, a video image recognition result collection module, a video image recognition sequence analysis module and a video analysis intercepting module which are sequentially connected.
The video image capturing module is used for capturing a video image according to a certain frequency, zooming the video image into a preset image and detecting the size of resolution; the video image recognition module is used for calling a video image recognition model which is trained in advance, obtaining the content of a specified single-frame video image through the video image capturing module, and sending the content into the video image recognition model to obtain a video image recognition result; the video image identification result collection module is used for initializing the length of the sequencer according to the original video length and the detection frequency and storing the video image identification result according to the time sequence; the video image identification sequence analysis module is used for completing detection and identification of the whole video according to a preset detection frequency, obtaining a detection result corresponding to each frame of image according to a time sequence, and traversing the sequencer to obtain time nodes corresponding to continuous types and types of corresponding videos; the video analyzing and intercepting module is used for intercepting a video according to the timestamp corresponding to the identification category sequence, acquiring video clips of the video and finally acquiring a plurality of video clips corresponding to different categories.
When the system works, video images are captured according to a certain frequency through the video image capturing module, 3-channel RGB images with fixed resolution are obtained through scaling processing, the images are sent to the video image recognition module to be recognized to obtain recognition results, a complete video stream is processed to obtain a video image recognition sequence and the video image recognition sequence is stored in the video image recognition result collection module, a plurality of video node information and corresponding category information are obtained through the video image recognition sequence analysis module, and finally the whole video is analyzed and processed through the video analysis capturing module to obtain a plurality of sub-video contents and corresponding labels, so that automatic video annotation and key sub-video content extraction are achieved.
The system can obtain a video image recognition result time sequence by recognizing the ordered video image content set, analyze the video image recognition result time sequence to obtain video key sub-video fragment node information and the fragment category information, obtain a plurality of key sub-videos and corresponding key sub-video category information according to the node information through the video analysis intercepting module, finally show the key sub-video content of the complete video, and label the complete video with a plurality of types of label information.
Specifically, in this embodiment, the video image recognition model is a convolutional neural network model. And the video image recognition model adopts a Mobile Net series network structure of google for a feature extraction network during feature extraction. The training process of the convolutional neural network model is as follows:
A. collecting image samples of specified resolutions of a plurality of categories, and setting network structure input; normalizing each collected image sample;
B. generating tfrecrd data format supported by tensorflow aiming at picture sample information and picture samples, and dividing the tfrecrd data format into a training file and a verification file according to a certain proportion, wherein the images of the training file and the verification file are different, and the stored image format is the same as the label information;
C. training the model by using a training file to generate a picture content classification model with appointed N classification contents, and verifying the picture content classification model by using a verification file;
D. and judging whether the model loss is reduced to a preset threshold or the training steps reach the preset steps, finishing the training if the model loss is reduced to the preset threshold or the training steps reach the preset steps, and increasing video content picture samples or adjusting model parameters and returning to the step A if the model loss is not reduced to the preset threshold and the training steps do not reach the preset steps until the training is finished.
Example two
A video classification intercepting method based on image recognition is realized by the video classification intercepting system based on image recognition in the embodiment I, and comprises the following steps:
s1, a video image capturing module captures a video image according to a certain frequency, zooms the video image into a preset picture and detects the size of resolution.
And S2, calling a video image recognition model trained in advance by the video image recognition module, obtaining the content of the appointed single-frame video image through the video image capturing module, and sending the content into the video image recognition model to obtain the video image recognition result.
Specifically, in this embodiment, the training process of the video image recognition model is as follows:
step 1, aiming at the input characteristics of the neural network and the preset category types, 5 million video image samples are respectively collected for different categories, and the 3-channel RGB images with the fixed resolution of 320 × 240 are uniformly set.
And 2, normalizing the RGB three-channel pixel values of the video picture to help the model to accelerate convergence. In this embodiment, the image is specifically normalized by the following formula:
Y=(X–127.5)/128。
and 3, generating tfrecrd data format supported by tensoflow aiming at the picture sample information and the picture sample, and dividing the tfrecrd data format into a training file and a verification file according to a certain proportion, wherein the images of the training file and the verification file are different, but the stored image format and the label information are the same.
And 4, training the model by using the training file to generate a video image recognition model, and verifying the picture content classification model by using the verification file.
Step 5, judging whether the model loss is reduced to a preset threshold (0.01 in the embodiment) or the training step number reaches a preset step number (5 ten thousand in the embodiment);
if the loss of the model is reduced to 0.01 of a preset threshold value or the number of training steps reaches 5 thousands of steps, finishing the training; otherwise, adding the video content picture sample or adjusting the model parameters (such as resetting the model loss threshold or the preset steps), and repeatedly executing the steps 1 to 5 until the training is completed.
And S3, initializing the length of the sequencer by the video image identification result collection module through the original video length and the detection frequency, and storing the video image identification result according to the time sequence.
In this embodiment, the video image recognition result collection module calculates the length of the sequencer according to the video duration and the detection frequency, and calculates the video image result at the specified frequency to obtain and store the result in the image recognition sequencer according to the time sequence until the video detection is completed or the sequencer is full.
S4, a video image identification sequence analysis module completes detection and identification of the whole video according to a preset detection frequency, obtains a detection result corresponding to each frame image according to a time sequence, and a traversal sequencer obtains time nodes corresponding to continuous types and types of corresponding videos;
the embodiment specifically includes: the video image recognition sequence analysis module traverses the image recognition sequencer, stores the current position N when recognizing the video type P, and records the current position M when recognizing the non-video type P, so that the current video clip is the sub-video type P from the time N to the time M; and finally outputting the starting time and the ending time of a plurality of video segments and marking corresponding class labels according to the mode until the image recognition sequencer is traversed. Fig. 2 is a schematic diagram illustrating analysis of a video image recognition sequence according to the present embodiment, where a start N indicates recognition of a start time of an image category P, an end M indicates recognition of an end time of the category P, and a black frame indicates a category P sub-video range.
S5, intercepting the video by a video analyzing and intercepting module according to the timestamp corresponding to the identification category sequence, acquiring video clips of the video, and finally acquiring a plurality of video clips corresponding to different categories;
the embodiment specifically includes: the video analyzing and intercepting module obtains time nodes corresponding to a plurality of image recognition sequencers, calculates the real time length of a corresponding video according to the video length and the detection frequency, cuts the complete video into a plurality of video segments, and prints a plurality of category labels on the complete video; and informing the user of the corresponding video category label through video recommendation analysis, and displaying the cut video segment to the user.
The video classification intercepting method based on image recognition can recognize various types of video information and extract and intercept the key sub-video content of the corresponding type, utilizes the image recognition technology to recognize and process a video image set, obtains the video type and video type node information by analyzing a recognition result sequence, finally obtains a plurality of key sub-video contents and a plurality of types of labels, can be effectively applied to the fields of video classification and video interception, realizes the automatic labeling of video files and the extraction of video highlights key videos, meets the requirements of the automatic video classification technology and the video intercepting method, and effectively improves the effectiveness and diversity of the extraction of the video content information.
It will be understood that the above embodiments are merely exemplary embodiments taken to illustrate the principles of the present invention, which is not limited thereto. It will be apparent to those skilled in the art that various modifications and improvements can be made without departing from the spirit and substance of the invention, and these modifications and improvements are also considered to be within the scope of the invention.
Claims (9)
1. A video classification intercepting system based on image recognition is characterized by comprising:
the video image capturing module is used for capturing a video image according to a certain frequency, zooming the video image into a preset image and detecting the size of resolution;
the video image recognition module is used for calling a video image recognition model which is trained in advance, obtaining the content of the appointed single-frame video image through the video image capturing module, and sending the content into the video image recognition model to obtain the video image recognition result;
the video image recognition result collection module is used for initializing the length of the sequencer according to the original video length and the detection frequency and storing the video image recognition result according to a time sequence;
the video image identification sequence analysis module is used for completing detection and identification of the whole video according to a preset detection frequency, obtaining a detection result corresponding to each frame of image according to a time sequence, and traversing the sequencer to obtain time nodes corresponding to continuous types and types of corresponding videos;
the video analyzing and intercepting module is used for intercepting a video according to the timestamp corresponding to the identification category sequence, acquiring video clips of the video and finally acquiring a plurality of video clips corresponding to different categories;
the video image capturing module, the video image recognition result collecting module, the video image recognition sequence analyzing module and the video analyzing and intercepting module are sequentially connected.
2. The image recognition-based video classification intercepting system of claim 1, wherein the video image recognition model is a convolutional neural network model.
3. The video classification intercepting system based on image recognition as claimed in claim 2, wherein the video image recognition model adopts a mobile net series network structure of google for the feature extraction network when performing feature extraction.
4. The video classification interception system based on image recognition according to claim 2, wherein the training procedure of the convolutional neural network model is as follows:
A. collecting image samples of specified resolutions of a plurality of categories, and setting network structure input;
B. generating tfrecrd data format supported by tensorflow aiming at picture sample information and picture samples, and dividing the tfrecrd data format into a training file and a verification file according to a certain proportion, wherein the images of the training file and the verification file are different, and the stored image format is the same as the label information;
C. training the model by using a training file to generate a picture content classification model with appointed N classification contents, and verifying the picture content classification model by using a verification file;
D. and judging whether the model loss is reduced to a preset threshold or the training steps reach the preset steps, finishing the training if the model loss is reduced to the preset threshold or the training steps reach the preset steps, and increasing video content picture samples or adjusting model parameters and returning to the step A if the model loss is not reduced to the preset threshold and the training steps do not reach the preset steps until the training is finished.
5. The image recognition-based video classification intercepting system of claim 4, wherein the step A further comprises performing normalization processing on each collected image sample.
6. A video classification intercepting method based on image recognition is realized by the video classification intercepting system based on image recognition of any one of claims 1 to 5, and is characterized by comprising the following steps:
s1, a video image capturing module captures a video image according to a certain frequency, zooms the video image into a preset picture and detects the size of resolution;
s2, calling a video image recognition model trained in advance by a video image recognition module, obtaining the content of a specified single-frame video image through a video image capturing module, and sending the content into the video image recognition model to obtain a video image recognition result;
s3, initializing the length of the sequencer by the video image identification result collection module through the original video length and the detection frequency, and storing the video image identification result according to the time sequence;
s4, a video image identification sequence analysis module completes detection and identification of the whole video according to a preset detection frequency, obtains a detection result corresponding to each frame image according to a time sequence, and a traversal sequencer obtains time nodes corresponding to continuous types and types of corresponding videos;
and S5, intercepting the video by the video analyzing and intercepting module according to the timestamp corresponding to the identification category sequence, acquiring the video clip of the video, and finally acquiring a plurality of video clips corresponding to different categories.
7. The method for video classification interception based on image recognition according to claim 6, wherein the step S3 specifically comprises:
the video image recognition result collection module calculates the length of the sequencer through the video time length and the detection frequency, and obtains and stores the result into the image recognition sequencer according to the time sequence by calculating the video image result under the appointed frequency until the video detection is finished or the sequencer is filled.
8. The method for video classification interception based on image recognition according to claim 6, wherein the step S4 specifically comprises:
the video image recognition sequence analysis module traverses the image recognition sequencer, stores the current position N when recognizing the video type P, and records the current position M when recognizing the non-video type P, so that the current video clip is the sub-video type P from the time N to the time M;
and finally outputting the starting time and the ending time of a plurality of video segments and marking corresponding class labels according to the mode until the image recognition sequencer is traversed.
9. The method for video classification interception based on image recognition according to claim 6, wherein the step S5 specifically comprises:
the video analyzing and intercepting module obtains time nodes corresponding to a plurality of image recognition sequencers, calculates the real time length of a corresponding video according to the video length and the detection frequency, cuts the complete video into a plurality of video segments, and prints a plurality of category labels on the complete video;
and informing the user of the corresponding video category label through video recommendation analysis, and displaying the cut video segment to the user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011247468.2A CN112380954A (en) | 2020-11-10 | 2020-11-10 | Video classification intercepting system and method based on image recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011247468.2A CN112380954A (en) | 2020-11-10 | 2020-11-10 | Video classification intercepting system and method based on image recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112380954A true CN112380954A (en) | 2021-02-19 |
Family
ID=74578243
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011247468.2A Pending CN112380954A (en) | 2020-11-10 | 2020-11-10 | Video classification intercepting system and method based on image recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112380954A (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693739A (en) * | 2011-03-24 | 2012-09-26 | 腾讯科技(深圳)有限公司 | Method and system for video clip generation |
CN103475935A (en) * | 2013-09-06 | 2013-12-25 | 北京锐安科技有限公司 | Method and device for retrieving video segments |
CN106897714A (en) * | 2017-03-23 | 2017-06-27 | 北京大学深圳研究生院 | A kind of video actions detection method based on convolutional neural networks |
CN108615043A (en) * | 2016-12-12 | 2018-10-02 | 中移(杭州)信息技术有限公司 | A kind of video classification methods and system |
CN109858514A (en) * | 2018-12-20 | 2019-06-07 | 北京以萨技术股份有限公司 | A kind of video behavior classification method neural network based |
CN109948446A (en) * | 2019-02-20 | 2019-06-28 | 北京奇艺世纪科技有限公司 | A kind of video clip processing method, device and computer readable storage medium |
CN110147700A (en) * | 2018-05-18 | 2019-08-20 | 腾讯科技(深圳)有限公司 | Video classification methods, device, storage medium and equipment |
CN110490112A (en) * | 2019-08-13 | 2019-11-22 | 新华智云科技有限公司 | Football video segment detection method, device, system and storage medium |
CN110751030A (en) * | 2019-09-12 | 2020-02-04 | 厦门网宿有限公司 | Video classification method, device and system |
CN111523566A (en) * | 2020-03-31 | 2020-08-11 | 易视腾科技股份有限公司 | Target video clip positioning method and device |
CN111753790A (en) * | 2020-07-01 | 2020-10-09 | 武汉楚精灵医疗科技有限公司 | Video classification method based on random forest algorithm |
-
2020
- 2020-11-10 CN CN202011247468.2A patent/CN112380954A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693739A (en) * | 2011-03-24 | 2012-09-26 | 腾讯科技(深圳)有限公司 | Method and system for video clip generation |
CN103475935A (en) * | 2013-09-06 | 2013-12-25 | 北京锐安科技有限公司 | Method and device for retrieving video segments |
CN108615043A (en) * | 2016-12-12 | 2018-10-02 | 中移(杭州)信息技术有限公司 | A kind of video classification methods and system |
CN106897714A (en) * | 2017-03-23 | 2017-06-27 | 北京大学深圳研究生院 | A kind of video actions detection method based on convolutional neural networks |
CN110147700A (en) * | 2018-05-18 | 2019-08-20 | 腾讯科技(深圳)有限公司 | Video classification methods, device, storage medium and equipment |
CN109858514A (en) * | 2018-12-20 | 2019-06-07 | 北京以萨技术股份有限公司 | A kind of video behavior classification method neural network based |
CN109948446A (en) * | 2019-02-20 | 2019-06-28 | 北京奇艺世纪科技有限公司 | A kind of video clip processing method, device and computer readable storage medium |
CN110490112A (en) * | 2019-08-13 | 2019-11-22 | 新华智云科技有限公司 | Football video segment detection method, device, system and storage medium |
CN110751030A (en) * | 2019-09-12 | 2020-02-04 | 厦门网宿有限公司 | Video classification method, device and system |
CN111523566A (en) * | 2020-03-31 | 2020-08-11 | 易视腾科技股份有限公司 | Target video clip positioning method and device |
CN111753790A (en) * | 2020-07-01 | 2020-10-09 | 武汉楚精灵医疗科技有限公司 | Video classification method based on random forest algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110008797B (en) | Multi-camera multi-face video continuous acquisition method | |
CN109284729B (en) | Method, device and medium for acquiring face recognition model training data based on video | |
CN107223246B (en) | Image labeling method and device and electronic equipment | |
CN110267061B (en) | News splitting method and system | |
Yang et al. | Lecture video indexing and analysis using video ocr technology | |
CN113365147B (en) | Video editing method, device, equipment and storage medium based on music card point | |
CN107180056B (en) | Method and device for matching segments in video | |
US20220172476A1 (en) | Video similarity detection method, apparatus, and device | |
CN115994230A (en) | Intelligent archive construction method integrating artificial intelligence and knowledge graph technology | |
CN110610123A (en) | Multi-target vehicle detection method and device, electronic equipment and storage medium | |
CN113065568A (en) | Target detection, attribute identification and tracking method and system | |
CN111428087B (en) | Video interception method and device, computer equipment and storage medium | |
WO2022062028A1 (en) | Wine label recognition method, wine information management method and apparatus, device, and storage medium | |
WO2022062027A1 (en) | Wine product positioning method and apparatus, wine product information management method and apparatus, and device and storage medium | |
CN114051154A (en) | News video strip splitting method and system | |
CN113537088A (en) | Digital image sensor hardware fingerprint matching method and system based on small sample learning | |
EP2345978A1 (en) | Detection of flash illuminated scenes in video clips and related ranking of video clips | |
CN112380954A (en) | Video classification intercepting system and method based on image recognition | |
CN116469167A (en) | Method and system for obtaining character action fragments based on character actions in video | |
CN113536032A (en) | Video sequence information mining system, method and application thereof | |
CN113689585B (en) | Non-inductive attendance card punching method, system and related equipment | |
CN111813996B (en) | Video searching method based on sampling parallelism of single frame and continuous multi-frame | |
CN114022905A (en) | Attribute-aware domain expansion pedestrian re-identification method and system | |
CN114245232A (en) | Video abstract generation method and device, storage medium and electronic equipment | |
CN112418215A (en) | Video classification identification method and device, storage medium and equipment |
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: 20210219 |