CN115908503A - Game video ROI detection method and system - Google Patents
Game video ROI detection method and system Download PDFInfo
- Publication number
- CN115908503A CN115908503A CN202310046077.1A CN202310046077A CN115908503A CN 115908503 A CN115908503 A CN 115908503A CN 202310046077 A CN202310046077 A CN 202310046077A CN 115908503 A CN115908503 A CN 115908503A
- Authority
- CN
- China
- Prior art keywords
- game video
- foreground
- optical flow
- game
- roi
- 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.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a game video ROI detection method and a system, belonging to the technical field of image recognition; wherein the method comprises the following steps: acquiring a first game video clip; performing first processing on the first game video segment to obtain a significant movement target; second processing the first game video segment to obtain an optical flow map of the first game video segment, determining a non-salient moving object from the optical flow map; and taking the significant moving target and the non-significant moving target as ROI detection result objects. The scheme of the invention does not depend on a ROI detection method based on data set labeling, thereby not only reducing the ROI detection cost, but also obviously improving the accuracy and comprehensiveness of ROI detection.
Description
Technical Field
The invention relates to the technical field of games and image recognition, in particular to a game video ROI detection method and system.
Background
The current ROI (Region of Interest) detection technology mainly has two main categories:
1) Detection techniques based on object detection or instance segmentation: this kind of approach usually considers the detection speed, and for the ROI detection task, a yolo series target detection network is often used as the recognition of foreground target in image or video frame, such as YOLOv3. The method mainly comprises the following steps: according to a specific image or video data set, target units of interest in a video are specified manually, various marking tools are used for marking the target units, a training data set is made, various deep learning frames are used for training the data set to obtain optimal model parameters, a trained model is used for reasoning unknown data, and the target units which are the same as or similar to the data set are detected and identified.
2) The method based on the significance target detection comprises the following steps: this type of method usually uses computer vision to detect salient objects in an image by simulating the human visual attention mechanism, and filters redundant background information, so as to focus only on the image area interested by human vision. The method mainly comprises the following steps: the method comprises the steps of marking interested units in an image or video frame by means of an eye tracker or artificial subjective judgment and the like, and carrying out segment-wise pixel-level modeling by a CNN-based full convolution network or an MLP-CNN fused hybrid network.
Through the analysis, the existing technology is a strong supervised learning mode, and is seriously dependent on the labeling of the data set, and the organization of the data set depends on the subjective judgment of people to a certain extent, so that the large-scale data set production is difficult. In addition, the game video is different from a daily life scene, the scene content of the game video is different from daily things, the games are various, data set production is carried out aiming at each game, and both time cost and labor cost cannot be borne.
Disclosure of Invention
In order to solve at least the technical problems in the background art, the invention provides a game video ROI detection method, a game video ROI detection system, electronic equipment and a computer storage medium.
The invention provides a game video ROI detection method, which comprises the following steps:
acquiring a first game video clip;
performing first processing on the first game video segment to obtain a significant moving object;
performing second processing on the first game video segment to obtain an optical flow diagram of the first game video segment, and determining a non-significant moving object according to the optical flow diagram;
and taking the significant movement target and the non-significant movement target as ROI detection result objects.
Further, the method further comprises: receiving a second game video clip;
performing stability analysis on each video frame of the second game video clip;
and extracting a plurality of first game video clips from the second game video clip according to the stability analysis result.
Further, the performing stability analysis on each video frame of the second game video segment includes:
extracting first background features of each video frame, and determining a first judgment threshold according to the first background features;
extracting second background features of each video frame, and determining a stability value according to the second background features;
and performing stability analysis according to the stability value and the first judgment threshold value to obtain the stability analysis result.
Further, the first processing the first game video segment to obtain a significant motion target includes:
identifying foreground objects in video frames of the first game video clip;
extracting a first foreground feature and a second foreground feature of each foreground object in adjacent video frames;
and identifying a plurality of the significant moving targets from the foreground targets according to the first foreground features, the second foreground features and a second judgment threshold.
Further, the determining non-salient moving objects from the optical flow graph includes:
determining optical flow characteristics according to the optical flow graph, and determining the optical flow characteristics of the foreground object which does not belong to the significant moving object;
and determining the foreground target corresponding to the optical flow characteristics meeting the preset conditions as the non-significant target.
Further, the first foreground feature, the second foreground feature and the optical flow feature are extracted by:
inputting video frames of the first game video clip and the optical flow graph into a downsampled Encoder network sharing parameters;
the down-sampling Encoder network outputs original features and the optical flow features; wherein the raw features comprise the first foreground feature and the second foreground feature.
Further, the method further comprises:
and receiving a third game video segment, and labeling the third game video segment according to the ROI detection result object.
The invention provides a game video ROI detection system, which comprises an acquisition module, a processing module and a storage module, wherein the processing module is connected with the acquisition module and the storage module; wherein, the first and the second end of the pipe are connected with each other,
the storage module is used for storing executable computer program codes;
the acquisition module is used for acquiring the game video clip and transmitting the game video clip to the processing module;
the processing module is configured to execute the method according to any one of the preceding claims by calling the executable computer program code in the storage module.
A third aspect of the present invention provides an electronic device comprising: a memory storing executable program code; a processor coupled with the memory; the processor calls the executable program code stored in the memory to perform the method of any of the preceding claims.
A fourth aspect of the invention provides a computer storage medium having stored thereon a computer program which, when executed by a processor, performs a method as set forth in any one of the preceding claims.
The technical scheme of the invention has the beneficial effects that:
1) The technical scheme provided by the invention does not depend on the labeling of a specific data set, and the training of the model can be completed only by using a conventional open source data set, so that the economic benefit is better;
2) The technical scheme provided by the invention does not depend on a specific game scene, has a wide application range, covers most popular games in the market and has very strong generalization capability;
3) Compared with the existing method, the method has the advantages that the game scene ROI detection result is better and more stable.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
FIG. 1 is a flow chart of a method for detecting ROI in a game video according to an embodiment of the invention.
FIG. 2 is a schematic diagram of foreground feature and optical flow feature extraction by using a downsampling Encoder network with shared parameters according to an embodiment of the present invention.
FIG. 3 is a schematic structural diagram of a system for detecting ROI in game video according to an embodiment of the present invention.
Fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms first, second, third and the like in the description and in the claims, or modules a, B, C and the like are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order, it being understood that specific orders or sequences may be interchanged, if appropriate, to enable embodiments of the invention described herein to be practiced otherwise than as specifically illustrated or described herein.
In the following description, reference numerals indicating steps such as S110, S120 \ 8230 \8230 \ 8230, etc. do not necessarily indicate that the steps are performed, and the order of the front and rear steps may be interchanged or performed simultaneously, where the case allows.
The term "comprising" as used in the specification and claims should not be construed as being limited to the contents listed thereafter; it does not exclude other elements or steps. It should therefore be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, and groups thereof. Thus, the expression "an apparatus comprising the devices a and B" should not be limited to an apparatus consisting of only the components a and B.
Reference in the specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. Thus, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments, as would be apparent to one of ordinary skill in the art from this disclosure.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. In the case of inconsistency, the meaning described in the present specification or the meaning derived from the content described in the present specification shall control. In addition, the terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to be limiting of the invention.
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for detecting ROI in game video according to an embodiment of the present invention. The game video ROI detection method provided by the embodiment of the invention comprises the following steps:
acquiring a first game video clip;
performing first processing on the first game video segment to obtain a significant movement target;
second processing the first game video segment to obtain an optical flow map of the first game video segment, determining a non-salient moving object from the optical flow map;
and taking the significant movement target and the non-significant movement target as ROI detection result objects.
In the embodiment of the invention, the ROI detection method in the prior art firstly trains the foreground recognition model by using the video frame of a specific game, however, manual marking of the video frame needs to consume a large amount of manpower, and the foreground recognition model needs to be trained again when ROI detection is carried out on different games, so that the prior ROI detection method is very inefficient due to the above main factors and is difficult to meet actual requirements.
Aiming at the technical problem, the method firstly identifies the significant moving target of the game video clip, and then carries out the complementary identification on the non-significant moving target through the optical flow analysis, thereby obtaining more comprehensive ROI detection effect. In the technical scheme of the invention, the model based on the first processing does not need to be pre-trained as in the prior art or only needs to be universally trained, and the second processing does not need to be trained at all, so that compared with the prior art, the technical scheme of the invention not only reduces the ROI detection cost, but also obviously improves the accuracy and comprehensiveness of ROI detection.
The prominent moving object may be "hero" with rapid movement in the glory of the royal, and the non-prominent moving object may be "wild" with slow movement, although the description is merely illustrative and not limiting the scope of the present invention.
Further, the method further comprises: receiving a second game video clip;
performing stability analysis on each video frame of the second game video clip;
and extracting a plurality of first game video clips from the second game video clip according to the stability analysis result.
In this step, when the game player moves the subject character (i.e., the character manipulated by the game player) to change the position in the game scene (e.g., during the running of the subject character), the game scene changes with the motion, and the background in the screen also changes greatly, which reduces the accuracy of the first processing result. Aiming at the problem, the invention carries out stability analysis on the background characteristics among the continuous video frames, selects a video clip with almost unchanged background, namely a first game video clip, and carries out ROI detection on the first game video clip. Since the background hardly changes, the difficulty of the first process can be significantly reduced, and accuracy can also be ensured.
The stability analysis may be implemented by analyzing a change of the background, such as an average value, a variance, and an accumulated value of a change of the background pixel, and/or a change rate, which is not limited by the present invention. And, the extraction of background features belongs to the mature technology in the field of image recognition, and the invention is not described herein in too much detail.
Further, the performing stability analysis on each video frame of the second game video segment includes:
extracting first background features of each video frame, and determining a first judgment threshold according to the first background features;
extracting second background features of each video frame, and determining a stability value according to the second background features;
and performing stability analysis according to the stability value and the first judgment threshold value to obtain the stability analysis result.
In this step, the determination threshold is a key content for performing the stability analysis, but it is actually difficult to accurately set the threshold for each frame of the video. Then, the present invention determines the depth of field of each video frame according to the first background feature of the video frame (close shot/far shot, zooming in/out of the picture), and determines the first decision threshold of the video frame accordingly; then, the stability analysis result is obtained by comparing the value with the stability value calculated according to the second background feature, for example, if the stability value is greater than or equal to the first determination threshold, the stability analysis result is determined to be stable.
The first background feature may be a background feature of a specific area, the main character is generally located at a substantially central position of the game screen, and the trees, mountains, etc. are generally located at the upper half of the screen. Then, a tree, a distant mountain and the like can be used as a specific area, the first background feature can be the size, the image definition and the like of a connected area corresponding to the tree and the distant mountain, and the smaller the size and/or the higher the definition are, the closer the depth of field is (namely, a close shot/a zoomed shot), the larger the first determination threshold is determined to be; otherwise, the farther the depth of field is (i.e., long shot/short shot), the smaller the first decision threshold. And, the main character may be defined as the specific region, and the first background feature may be a moving speed (i.e., an amount of displacement in the game screen per unit time) of the main character, and the faster the moving speed, the closer the depth, the larger the first determination threshold; otherwise, the farther the depth of field is, the smaller the first determination threshold is, and the correspondence between the specific depth of field and the first determination threshold may be set. By the method, the first judgment threshold values of different video frames can be determined quickly and pertinently. Wherein the subject character may be identified based on a particular characteristic, such as an identifying halo (halo located under the subject character's foot) for a game player to identify the subject character.
The second background feature is a background feature for the whole video frame, for example, an equivalent pixel value of a whole background pixel of each video frame, or an equivalent pixel value of a pixel in an edge region of each video frame.
In the embodiment of the invention, the stability analysis of the whole video frame about the background is carried out, and the background extraction is not needed in the process, so that the processing load can be effectively reduced, and the analysis efficiency can be improved. Moreover, by increasing the first judgment threshold of the close shot, the influence of the movement of the main character with a larger area in the game picture on the second background characteristic can be weakened, and the stability analysis accuracy of each video frame is further improved.
Further, the first processing the first game video segment to obtain a significant motion target includes:
identifying foreground objects in video frames of the first game video clip;
extracting a first foreground feature and a second foreground feature of each foreground object in adjacent video frames;
and identifying a plurality of the significant moving targets from the foreground targets according to the first foreground features, the second foreground features and a second judgment threshold.
In this step, a significant moving object may be determined by a feature difference of each foreground object in adjacent video frames, for example, the first/second foreground features may be positions, pixel changes, and the like of the foreground object, and when a position change (fast movement) or a pixel change (change of orientation, station/sitting posture switching, and the like) of the foreground object in the adjacent frames meets a second determination threshold occurs, the foreground object may be identified as a significant moving object.
Further, the determining non-salient moving objects from the optical flow graph includes:
determining optical flow characteristics according to the optical flow graph, and determining the optical flow characteristics of the foreground object which does not belong to the significant moving object;
and determining the foreground target corresponding to the optical flow features meeting preset conditions as the non-significant target.
In this step, when determining the significant moving object, the second determination threshold is generally set to be larger to avoid erroneous determination, which results in that part of the foreground object is not recognized. In response to the problem, the invention further re-identifies the features through optical flow features, and further can screen out non-significant objects. The obtaining method of the light flow graph may adopt a mature scheme in the prior art, and the present invention is not limited to this and is not described in detail herein.
Further, the first foreground feature, the second foreground feature, and the optical flow feature are extracted by:
inputting video frames of the first game video clip and the optical flow graph into a downsampled Encoder network sharing parameters;
the downsampling Encoder network outputs original features and the optical flow features; wherein the original features include the first foreground feature and the second foreground feature.
In this step, as shown in fig. 2, in the moving object detection stage, each video frame in the original first game video segment and the optical flow graph obtained in the foregoing are input to an Encoder downsampling network sharing one parameter to perform feature extraction, and an original feature and an optical flow feature are obtained respectively.
In addition, before the above features are analyzed (i.e. before the features are input into a Decoder), the above two features need to be fused, the original features are used for detecting foreground objects with significant motion, the optical flow features are used for providing strong supervision signals for foreground objects with non-significant motion, after feature fusion, the binary mask output result of restoring the original resolution by utilizing upsampling is the output of the whole process, that is, the significant motion object and the non-significant motion object are taken as ROI detection result objects.
In this case, optical flow diagrams of consecutive frames (for example, 5 frames) may be simultaneously input or a downsampling Encoder network may be input in units of groups, so that accuracy of optical flow feature extraction may be improved.
Further, the method further comprises:
and receiving a third game video segment, and labeling the third game video segment according to the ROI detection result object.
In this step, after each foreground object, i.e. ROI detection result object, is accurately identified, it may be tracked to realize labeling in the subsequent third game video segment. The third game video segment may be a historical game video segment acquired simultaneously with the first game video segment and the second game video segment, or may be accessed real-time video stream data, that is, a real-time game video.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a game video ROI detection system according to an embodiment of the present invention. As shown in fig. 3, a system for detecting ROI of game video according to an embodiment of the present invention includes an obtaining module 101, a processing module 102, and a storage module 103, where the processing module 102 is connected to the obtaining module 101 and the storage module 103; wherein the content of the first and second substances,
the storage module 103 is used for storing executable computer program codes;
the acquisition module 101 is configured to acquire a game video clip and transmit the game video clip to the processing module 102;
the processing module 103 is configured to execute the method according to the first embodiment by calling the executable computer program code in the storage module 103.
For specific functions of the game video ROI detection system in this embodiment, reference is made to the first embodiment, and since the system in this embodiment adopts all technical solutions of the first embodiment, at least all beneficial effects brought by the technical solutions of the first embodiment are achieved, and details are not repeated here.
Referring to fig. 4, fig. 4 is an electronic device according to an embodiment of the present invention, including: a memory storing executable program code; a processor coupled with the memory; the processor calls the executable program code stored in the memory to execute the method according to the embodiment.
The embodiment of the invention also discloses a computer storage medium, wherein a computer program is stored on the storage medium, and the computer program executes the method in the first embodiment when being executed by a processor.
Computer storage media for embodiments of the present invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
It should be noted that the foregoing is only a preferred embodiment 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 as will now become apparent to those skilled in the art 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.
Claims (10)
1. A game video ROI detection method is characterized by comprising the following steps:
acquiring a first game video clip;
performing first processing on the first game video segment to obtain a significant movement target;
second processing the first game video segment to obtain an optical flow map of the first game video segment, determining a non-salient moving object from the optical flow map;
and taking the significant movement target and the non-significant movement target as ROI detection result objects.
2. The method of claim 1, wherein the ROI is selected from a group consisting of: the method further comprises the following steps:
receiving a second game video clip;
performing stability analysis on each video frame of the second game video clip;
and extracting a plurality of first game video clips from the second game video clip according to the stability analysis result.
3. The method of claim 2, wherein the ROI is a video of a game: the performing stability analysis on each video frame of the second game video clip comprises:
extracting first background features of each video frame, and determining a first judgment threshold according to the first background features;
extracting second background features of each video frame, and determining a stability value according to the second background features;
and performing stability analysis according to the stability value and the first judgment threshold value to obtain the stability analysis result.
4. The method of claim 3, wherein the ROI of the game video is detected by: the first processing the first game video segment to obtain a salient moving object comprises:
identifying foreground objects in video frames of the first game video clip;
extracting a first foreground feature and a second foreground feature of each foreground object in adjacent video frames;
and identifying a plurality of the significant moving targets from the foreground targets according to the first foreground features, the second foreground features and a second judgment threshold.
5. The method of claim 4, wherein the ROI of the game video is detected by: said determining non-salient moving objects from said optical flow graph comprising:
determining optical flow characteristics according to the optical flow graph, and determining the optical flow characteristics of the foreground object which does not belong to the significant moving object;
and determining the foreground target corresponding to the optical flow features meeting preset conditions as the non-significant target.
6. The method of claim 4, wherein the ROI of the game video is detected by: the first foreground feature, the second foreground feature, and the optical flow feature are extracted by:
inputting each video frame of the first game video clip and the optical flow graph into a downsampled Encoder network of shared parameters;
the downsampling Encoder network outputs original features and the optical flow features; wherein the original features include the first foreground feature and the second foreground feature.
7. The method of ROI detection in game video according to any one of claims 1-6, wherein: the method further comprises the following steps:
and receiving a third game video clip, and labeling the third game video clip according to the ROI detection result object.
8. A game video ROI detection system comprises an acquisition module, a processing module and a storage module, wherein the processing module is connected with the acquisition module and the storage module; wherein the content of the first and second substances,
the storage module is used for storing executable computer program codes;
the acquisition module is used for acquiring a game video clip and transmitting the game video clip to the processing module;
the method is characterized in that: the processing module for performing the method of any one of claims 1-7 by invoking the executable computer program code in the storage module.
9. An electronic device, comprising:
a memory storing executable program code;
a processor coupled with the memory;
the method is characterized in that: the processor calls the executable program code stored in the memory to perform the method of any of claims 1-7.
10. A computer storage medium having a computer program stored thereon, characterized in that: the computer program, when executed by a processor, performs the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310046077.1A CN115908503B (en) | 2023-01-30 | 2023-01-30 | Game video ROI detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310046077.1A CN115908503B (en) | 2023-01-30 | 2023-01-30 | Game video ROI detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115908503A true CN115908503A (en) | 2023-04-04 |
| CN115908503B CN115908503B (en) | 2023-07-14 |
Family
ID=86491585
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310046077.1A Active CN115908503B (en) | 2023-01-30 | 2023-01-30 | Game video ROI detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115908503B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108810538A (en) * | 2018-06-08 | 2018-11-13 | 腾讯科技（深圳）有限公司 | Method for video coding, device, terminal and storage medium |
| CN111447449A (en) * | 2020-04-01 | 2020-07-24 | 北京奥维视讯科技有限责任公司 | ROI-based video coding method and system and video transmission and coding system |
| CN111586413A (en) * | 2020-06-05 | 2020-08-25 | 广州繁星互娱信息科技有限公司 | Video adjusting method and device, computer equipment and storage medium |
-
2023
- 2023-01-30 CN CN202310046077.1A patent/CN115908503B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108810538A (en) * | 2018-06-08 | 2018-11-13 | 腾讯科技（深圳）有限公司 | Method for video coding, device, terminal and storage medium |
| CN111447449A (en) * | 2020-04-01 | 2020-07-24 | 北京奥维视讯科技有限责任公司 | ROI-based video coding method and system and video transmission and coding system |
| CN111586413A (en) * | 2020-06-05 | 2020-08-25 | 广州繁星互娱信息科技有限公司 | Video adjusting method and device, computer equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| 徐金婧;徐熙平;: "基于时空显著特征的运动目标检测方法研究", 长春理工大学学报(自然科学版), no. 04, pages 43 - 48 * |
| 黄锋 等: "光流法和显著性相结合的动态背景下运动目标检测方法", 《成都工业学院学报》, no. 01, pages 13 - 18 * |
| 黄锋;易嘉闻;吴健辉;何伟;李武劲;欧先锋;: "光流法和显著性相结合的动态背景下运动目标检测方法", 成都工业学院学报, no. 01, pages 13 - 18 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115908503B (en) | 2023-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108846365B (en) | Detection method and device for fighting behavior in video, storage medium and processor | |
| CN110610510B (en) | Target tracking method and device, electronic equipment and storage medium | |
| WO2017000466A1 (en) | Method and system for tracking moving target based on optical flow method | |
| CN110795595A (en) | Video structured storage method, device, equipment and medium based on edge calculation | |
| EP4053735A1 (en) | Method for structuring pedestrian information, device, apparatus and storage medium | |
| CN102034240A (en) | Method for detecting and tracking static foreground | |
| CN112488071B (en) | Method, device, electronic equipment and storage medium for extracting pedestrian features | |
| CN113128368B (en) | Method, device and system for detecting character interaction relationship | |
| CN103093198A (en) | Crowd density monitoring method and device | |
| CN114140832A (en) | Method and device for detecting pedestrian boundary crossing risk in well, electronic equipment and storage medium | |
| CN111161325A (en) | Three-dimensional multi-target tracking method based on Kalman filtering and LSTM | |
| CN113515536A (en) | Map updating method, device, equipment, server and storage medium | |
| CN112560579A (en) | Obstacle detection method based on artificial intelligence | |
| CN114549557A (en) | Portrait segmentation network training method, device, equipment and medium | |
| CN109002808B (en) | Human behavior recognition method and system | |
| CN112949451B (en) | Cross-modal target tracking method and system through modal perception feature learning | |
| CN113139431B (en) | Image saliency target detection method based on deep supervised learning | |
| CN112866654B (en) | Intelligent video monitoring system | |
| CN110598540A (en) | Method and system for extracting gait contour map in monitoring video | |
| CN113688804A (en) | Multi-angle video-based action identification method and related equipment | |
| CN112422895A (en) | Image analysis tracking and positioning system and method based on unmanned aerial vehicle | |
| CN115908503A (en) | Game video ROI detection method and system | |
| CN115713731A (en) | Crowd scene pedestrian detection model construction method and crowd scene pedestrian detection method | |
| CN112200840B (en) | Moving object detection system in visible light and infrared image combination | |
| CN114066841A (en) | Sky detection method and device, computer equipment and 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |