Moving Body Detection Using Template Superposition and Candidate Position Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving body detection systems face difficulties in accurately tracking objects with few externally outstanding features, such as a white ball in a ball game, when the object's image overlaps with the background, sharing similar coloring and shape, leading to potential loss of target detection.
Innovation Solution
A moving body detecting device that predicts candidate positions, generates composite images by superimposing a template image of the target on the background image, and calculates a degree of match to accurately detect the target moving body, employing a candidate position predicting unit, composite image generating unit, degree-of-match calculating unit, and moving body detecting unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If external appearance features (color, shape) are used to detect the moving body, then the detection process is simple, but the detection accuracy deteriorates when the moving body has few externally outstanding features and overlaps with similar background
Solution Approach 1:
The system performs preliminary action by predicting candidate positions of the moving body before actual detection. The candidate position predicting unit uses motion models and previous frame data to anticipate where the moving body is likely to be, then focuses detection resources only on these predicted regions rather than scanning the entire image, thus maintaining simplicity while improving accuracy.
Solution Approach 2:
The system applies local quality by using template matching specifically at predicted candidate positions rather than throughout the entire image. The template image comparing unit creates a composite image by superimposing the template on the background at each candidate position and compares it with the actual image data only in those localized regions, improving detection accuracy for objects with few distinguishing features without significantly increasing overall system complexity.
2Measurement precision
If template matching is performed across the entire image to improve detection accuracy, then the detection accuracy improves, but the computational load and processing time increase significantly
Solution Approach 1:
The system segments the detection process into two distinct stages: first, the candidate position predicting unit identifies a limited set of likely positions using motion prediction; second, the template image comparing unit performs detailed template matching only at these segmented candidate positions. This segmentation reduces the search space from the entire image to specific regions of interest, maintaining high detection accuracy while significantly reducing computational load and processing time.
3Stability of the object's composition
If the moving body has similar coloring and shape to the background, then the blending is natural and less obtrusive, but the detection becomes difficult or impossible
Solution Approach 1:
The system uses an intermediary approach by introducing a template image as a reference model and creating a composite image that combines the template with the background. The template image comparing unit then compares this composite image with the actual image data at candidate positions. This intermediary template matching process enables detection of objects that blend with the background by looking for patterns consistent with the template rather than relying on color or shape differences.
Data Source
AI summary
In tracking of a moving body using image data in detecting a target moving body having few externally outstanding features, such as coloring and a shape, from an image in an image frame, if an image of the target moving body overlaps a background image largely and their colorings are the same or similar to one another, it is difficult to detect the target moving body. A template image of the target moving body is superimposed on a background image of a region (periphery) of a candidate position of the target to generate a composite image and the moving body is detected based on a degree of match calculated through comparison of the composite image with an image of the region (periphery) of the candidate position of the target.