Template-Based Subject Tracking with Target Region Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional subject tracking techniques often fail to accurately track the intended subject due to misalignment of the specified tracking target region, leading to incorrect template matching and loss of tracking.
Innovation Solution
An image processing apparatus and method that dynamically adjusts the tracking target region by generating a template based on the specified region and detecting similar regions in subsequent frames, using template matching to update the tracking target region to align with the intended subject.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If template matching is used to track a subject, then tracking can be performed based on image similarity, but the tracking fails when the initially specified tracking target region is misaligned with the intended subject
Solution Approach 1:
The patent implements feedback by detecting the subject region in each captured image and comparing it with the initially specified tracking target region. When a deviation is detected, the system automatically updates the tracking target region to align with the detected subject, creating a closed-loop control system that corrects initial specification errors and maintains reliable tracking throughout the sequence.
Solution Approach 2:
The patent performs preliminary subject detection and region alignment before initiating the tracking process. By pre-processing the initial frame to accurately identify the subject region and adjust the tracking target region accordingly, the system establishes a correct baseline for subsequent template matching operations, preventing tracking failures from the outset.
2Ease of operation
If the user specifies the tracking target by touch operation or pointer movement, then the user can select a desired position, but the specified position may shift from the intended subject due to handheld operation instability
Solution Approach 1:
The patent enables the system to automatically correct its own tracking target region without requiring precise manual specification. The subject detection and region update functions allow the system to self-adjust the tracking target based on actual subject position in the image, compensating for user specification errors and eliminating the need for high-precision manual input.
Solution Approach 2:
The system continuously monitors the detected subject region and compares it with the tracking target region, providing feedback that triggers automatic updates when misalignment is detected. This feedback mechanism compensates for initial positioning errors caused by handheld operation, maintaining accurate tracking despite imprecise user input.
3Reliability
If the tracking target region is updated based on detected similar regions, then accurate tracking can be maintained, but additional processing steps are required beyond simple template matching
Solution Approach 1:
The patent merges subject detection functionality with template matching tracking by integrating the region detection results directly into the tracking target update process. By combining these functions into a unified processing pipeline where detection informs tracking updates, the system achieves accurate tracking without requiring completely separate complex subsystems.
Solution Approach 2:
The patent implements dynamic updating of the tracking target region based on real-time subject detection results. The system adaptively adjusts the tracking target region in response to detected subject position and appearance changes, allowing the tracking parameters to evolve dynamically throughout the image sequence rather than remaining static.
Data Source
AI summary
An image processing apparatus includes one or more processors that execute a program and thereby function as a setting unit that sets a tracking target region, a generating unit that generates a template to be used in template matching based on a set tracking target region, and a detecting unit that detects, in a current image, a first region that is similar to a template generated by the generating unit, by applying template matching using the template to the image. Therein the setting unit sets, as a new tracking target region, the first region detected in the current image by the detecting unit or a second region of the current image whose position corresponds to a previously set tracking target region.