Object Classification via Image Area Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for object classification in sports videos, particularly in IPTV systems, face challenges in accurately distinguishing between players and referees with low computational complexity, especially when using regular broadcasting videos without fixed cameras, which are less flexible but more common.
Innovation Solution
A multi-histogram matching scheme combined with a playfield model-based segmentation method is proposed, where the image area around the object is divided into upper and lower portions using a discriminative function, enhancing classification accuracy while reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If color segmentation and morphological operations are used to find the playfield and limit the search area, then the classification accuracy is improved, but the computational complexity increases
Solution Approach 1:
The area around the located object is divided into at least a first portion and a second portion based on image information within the area. This segmentation allows the system to process only relevant portions of the image rather than the entire image, reducing computational complexity while maintaining classification accuracy.
Solution Approach 2:
Different portions of the image area are processed differently based on their relevance to the classification task. The method applies classification based on image information in specific portions rather than uniformly processing the entire image, optimizing computational resources while preserving accuracy.
2Measurement precision
If fixed cameras in controlled environments are used for locating players and referees, then the localization accuracy is improved, but the system flexibility and adaptability deteriorate
Solution Approach 1:
The method using regular broadcasting videos with playfield model-based segmentation can be applied to various sports and broadcasting scenarios without requiring fixed camera installations. This universal approach maintains adaptability across different environments while achieving acceptable localization accuracy through intelligent image processing.
Solution Approach 2:
The system adapts to different broadcasting conditions by adjusting processing parameters such as color space transformations, histogram matching thresholds, and playfield model parameters. This allows the system to maintain functionality across varying video qualities, lighting conditions, and camera angles without requiring fixed camera infrastructure.
3Measurement precision
If the entire image area is processed for object classification, then the classification completeness is improved, but the processing time and computational requirements increase
Solution Approach 1:
The image area around the located object is segmented into at least a first portion and a second portion, allowing the system to focus computational resources on relevant regions. This segmentation maintains classification completeness by examining all necessary portions while reducing the total processing area compared to analyzing the entire image.
Solution Approach 2:
The method processes only the necessary portions of the image area required for accurate classification rather than the entire image. This partial processing approach reduces computational time and resources while maintaining sufficient classification completeness for the application requirements.
Data Source
AI summary
In an image processing method, an object is located within an image. An area around the object is determined and divided into at least first and second portions based upon image information within the area. The object can then be classified based upon both image information in the first portion of the area and image information in the second portion of the area.