Target Tracking Apparatus Using Pixel Difference Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing photographing apparatuses face errors in target extraction and background identification when tracking moving targets, often incorrectly identifying moving objects as targets or failing to distinguish between targets and backgrounds.
Innovation Solution
A photographing apparatus and method that utilize a target extractor and tracking controller to generate and analyze target-difference and background-difference images, segmenting pixels based on pixel value differences and similarities to precisely extract and track targets by excluding background areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If target extraction is performed by analyzing captured images, then automatic tracking can be achieved, but extraction precision deteriorates due to errors in identifying targets versus background objects
Solution Approach 1:
The patent applies preliminary action by extracting and storing characteristic information (color, shape, size) of the target from the previous frame before processing the current frame. This pre-extracted information serves as a reference template that enables more accurate target identification in the current frame, reducing extraction errors while maintaining automatic tracking functionality.
Solution Approach 2:
The patent implements feedback by comparing the extracted target information from the current frame with the stored previous frame information. This comparison process provides feedback that corrects extraction errors and refines the target identification, thereby improving precision while preserving automated tracking operation.
2Quantity of substance
If background areas are included in target extraction, then extraction coverage increases, but tracking accuracy deteriorates due to inclusion of non-target objects
Solution Approach 1:
The patent applies segmentation by dividing the image into distinct target and background regions based on characteristic information comparison. The target extraction process segments only the regions that match the stored target characteristics (color, shape, size), thereby accurately separating targets from background objects while maintaining comprehensive target coverage.
Solution Approach 2:
The patent implements local quality by applying different extraction criteria to different regions of the image. Regions matching the stored target characteristics are extracted as targets, while regions with different characteristics are identified as background. This localized differentiation ensures high extraction accuracy while maintaining complete target area coverage.
Data Source
AI summary
A photographing apparatus and a method for controlling target tracking are provided. The photographing apparatus includes a target extractor for extracting target information in a present frame based on the difference between a pixel value of one pixel among previous pixels constituting the target of the previous frame in an area adjacent to the present pixel constituting the present frame and a pixel value of the present pixel, and a tracking controller for controlling automatic tracking for the target based on the target information. Accordingly, the target can be extracted more precisely.