Object Tracking With Backward-Verified Template Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object tracking systems face challenges in maintaining tracking performance due to changes in the appearance of target objects, often leading to inaccurate template updates and reduced efficiency.
Innovation Solution
The method employs a combination of forward and backward object tracking using a shared object tracking model to determine template candidates, evaluating their confidence based on overlap and historical tracking results, ensuring accurate template updates and improved tracking performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If template update is performed without verification, then tracking speed is improved, but tracking accuracy deteriorates due to appearance changes
Solution Approach 1:
The patent performs preliminary verification of template candidates through backward object tracking before actually updating the template. This preliminary action filters out incorrect templates that may cause tracking failures, ensuring only verified accurate templates are used for updates, thus maintaining both speed and accuracy.
Solution Approach 2:
The patent implements feedback by using the result of backward object tracking to evaluate and verify template candidates. The confidence score generated from comparing forward and backward tracking results provides feedback on template quality, allowing the system to select only verified accurate templates for updates.
2Measurement precision
If backward object tracking is performed to verify template candidates, then template accuracy is improved, but computational time increases
Solution Approach 1:
The patent applies partial action by performing backward object tracking only on a subset of candidate templates rather than all possible templates. This selective verification approach reduces computational time while maintaining sufficient accuracy by focusing verification resources on the most promising candidates.
3Adaptability or versatility
If template is updated frequently to adapt to appearance changes, then adaptability is improved, but tracking stability deteriorates due to inaccurate updates
Solution Approach 1:
The patent uses feedback from backward object tracking to verify template candidates before updates. This feedback mechanism ensures that only templates proven to accurately represent the target object are applied, maintaining tracking stability while still enabling frequent updates to adapt to appearance changes.
Solution Approach 2:
The patent performs preliminary verification of template candidates through backward tracking before applying updates. This preliminary action filters out unstable or inaccurate templates, ensuring that frequent updates to adapt to appearance changes do not compromise tracking stability.
Data Source
AI summary
A processor-implemented method with object tracking includes: performing, using a first template, forward object tracking on first image frames in a first sequence group; determining a template candidate of a second template for second image frames in a second sequence group; performing backward object tracking on the first image frames using the template candidate; determining a confidence of the template candidate using a result of comparing a first tracking result determined by the forward object tracking performed on the first image frames and a second tracking result determined by the backward object tracking performed on the first image frames; determining the second template based on the confidence of the template candidate; and performing forward object tracking on the second image frames using the second template.