Human Detection Tracking Using Absolute Frame Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human detection and tracking systems face challenges in accurately locating body parts when individuals are dressed in the same color, leading to errors in size and location due to the reliance on relative positional relationships between body part frames, which are affected by movements.
Innovation Solution
The system includes a human frame detecting section, a body part frame location determining section, and a correcting section that uses human feature data and body part feature data to detect and correct the location of body part frames, ensuring accurate tracking by referencing an absolute human frame.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If body part tracking is based only on relative positional relationship between body part frames, then the tracking method is simple, but the location accuracy of body parts deteriorates when individuals wear same-color clothing
Solution Approach 1:
The patent introduces a human frame as an intermediary reference object to resolve the tracking accuracy problem. Instead of directly tracking body parts using only relative relationships, the system uses the human frame (detected through human feature data) as a mediator to establish absolute positional relationships. The body part frame location correcting section then corrects body part locations based on the human frame's absolute position, thereby maintaining accuracy even when body parts have identical features.
2Productivity
If body part tracking relies on feature matching (luminance edge, color, texture), then the method is computationally efficient, but the tracking reliability deteriorates when body parts have identical features
Solution Approach 1:
The human frame serves as a reliable intermediary that does not suffer from the identical feature problem. Since the human frame is detected using comprehensive human feature data and represents the entire human body, it provides a stable reference that doesn't confuse body parts with same-color clothing. The correction section then uses this reliable reference to maintain tracking reliability without requiring complex re-identification of identical body parts.
Solution Approach 2:
The patent substitutes the mechanical feature-matching approach (which fails for identical features) with a coordinate transformation approach based on the human frame. Instead of relying on visual feature similarity to distinguish body parts, the system uses the human frame's absolute position and body part relative positions to calculate and correct body part locations, replacing feature-based identification with geometric relationship-based tracking.
3Speed
If the system uses only relative positional relationships for tracking, then the processing is fast, but the size and location errors increase between actual image data and tracking results
Solution Approach 1:
The system performs preliminary detection of the human frame using human feature data before conducting body part tracking. This preliminary action establishes the absolute reference framework that will be used for subsequent corrections. By preparing the human frame information in advance, the system can quickly correct body part locations without requiring complex real-time recalculations, thus maintaining processing speed while improving precision.
Solution Approach 2:
The patent replaces pure relative positional tracking with a hybrid system that incorporates absolute positioning through the human frame. The correction section uses the human frame's absolute coordinates to transform body part locations from relative to absolute reference, substituting the insufficient relative relationship mechanism with a more accurate absolute reference system that maintains both speed and precision.
Data Source
AI summary
A human detection and tracking apparatus prevents errors in a size of a person and a location of body parts between actual image data and a tracking result. Human frame detecting section detects, from first image data, a human frame as a region having high possibility of presence of a human, based on human feature data representing a feature of an entire human body. Body part frame location determining section determines a body part frame in the first image data, based on part feature data illustrating a feature of a body part of the human and a part frame determined as a region having high possibility of presence of a body part of the human in second image data previous to the first image data. Body part frame location correcting section corrects, based on the human frame, a location of the part frame determined in the first image data.