Multi-Object Tracking via Binary Frame Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-object detection and tracking systems face significant speed reductions when processing high-definition or high-frequency videos, as frame-by-frame detection methods are computationally intensive and fail to achieve real-time processing without compromising accuracy.
Innovation Solution
A tracking-by-detection system that employs binary search to divide frames into segments of steady motion, utilizing a preprocess unit to select sampled frames, a global object detector for whole-image detection, a local object detector for region-of-interest detection, and a data association unit to align detection results with existing trajectories, thereby reducing the number of frames that need to be detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If frame-by-frame detection is performed on the whole image, then detection accuracy is maintained, but processing speed becomes too slow for real-time applications
Solution Approach 1:
The patent divides the video sequence into segments based on motion coherence analysis. Instead of processing every frame, only key frames within each segment are detected, while other frames are handled through tracking. This segmentation approach maintains detection accuracy on important frames while dramatically reducing overall processing load to achieve real-time performance.
Solution Approach 2:
The patent applies detection partially - not on every frame, but selectively on key frames identified through motion analysis. By performing full detection only when necessary (on key frames) and using lighter tracking operations on intermediate frames, the system maintains accuracy where needed while improving overall processing speed.
2Measurement precision
If video resolution is increased to high definition, then image quality improves, but both detection and tracking speeds become slower
Solution Approach 1:
The patent segments high-definition video processing into detection phases (on key frames) and tracking phases (on intermediate frames). This allows the system to maintain high image quality for accurate detection on selected frames while reducing the computational burden by not performing full detection on every frame, thus achieving real-time processing of HD video.
Solution Approach 2:
The patent dynamically adjusts the processing intensity for each frame based on motion coherence analysis. Frames with significant motion changes receive full detection processing, while frames with coherent motion patterns use lighter tracking operations. This dynamic approach maintains image quality where needed while optimizing processing speed across the entire video sequence.
Data Source
AI summary
The present invention relates to the fast multi-object detection and tracking system. According to the system of the present invention, only a few frames are to be detected and the other frames are split into different size of steady motion segments in a binary way, and then the system of the present invention can predicted accurately. The system can help to achieve high tracking speed with multi-persons in high-definition videos and gain high accuracy as well.