Method and system for judging goals for basketball game video
A basketball and video technology, applied in the field of video goal determination in basketball games, can solve the problems of inability to handle real-time video streams and difficult training, and achieve the effect of reducing misjudgments and improving accuracy.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0065] Embodiment 1, a kind of method that is used for basketball match video goal judgment, as Figures 1 to 3 shown, including the following steps:
[0066] S1. Obtain all shooting moments in the video to be determined, and obtain corresponding shooting clips according to the shooting moments.
[0067] First, the detection model is obtained based on deep learning algorithm training, and each video frame in the video to be determined is detected by using the detection model, and the position, size and category (basketball / basket) of the basketball and the basket in each video frame are output, that is, the video frame The detection frame information of basketball and basket. According to the position, size and category (basketball / basketball) information of the basketball and basket in each video frame, it is judged whether there is a shot. If there is, the moment corresponding to the video frame is the moment of shooting. video clips to get shot clips.
[0068] The specif...
Embodiment 2
[0113] Embodiment 2, change goal data in the embodiment 1 step S2 from "goal, no goal" to "0 (i.e. no goal), 1 point (free throw), 2 points, 3 points", Be about to change the scoring model of the two classifications of embodiment 1 judgment goal / no goal into the scoring model of the four classifications of judging no goal, 1 point (free throw), 2 points, and 3 points, and all the other are equal to the implementation example 1.
[0114] The details of constructing the scoring model are as follows:
[0115] Manually mark all the sample shooting clips, and the tag content is the goal data of the video clip. The goal data includes 0 (no goal), 1 point (free throw), 2 points, 3 points, and the sample shooting clips and their Goal data. And the sample shot segment and its goal data are randomly divided into a detection video training set, a video verification set and a video test set according to Example 1.
[0116] The two-stream network TSN (Temporal Segment Networks, time dom...
Embodiment 3
[0121] Embodiment 3, a kind of system that is used for basketball game video goal determination, such as Figure 4 Shown include:
[0122] The game image acquisition module 1 is used to acquire the game image and its label information, and divide the game image and its label information into an image training set, an image verification set and an image test set according to a preset ratio. The label information includes basketball and basket the position, size and type of;
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com