Multi-Camera Target Tracking With Overlap-Based ID Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-target tracking systems face inaccuracies due to differing appearance features of objects in images collected by different devices, leading to incorrect tracking identifier assignments.
Innovation Solution
A method that synchronously acquires images from multiple devices, maps detection blocks using a mapping relation, and fuses target objects based on intersection over union (IOU) and appearance feature similarity to ensure consistent tracking identifiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If appearance feature extraction is performed independently on images from different image-collecting devices, then the tracking system can process images from multiple devices, but the appearance features will differ due to different collecting views, causing matching failure and inaccurate tracking
Solution Approach 1:
The patent introduces a mapping relation as an intermediary between images from different image-collecting devices. This mapping relation transforms appearance features from one device's coordinate system to another device's coordinate system, enabling consistent feature matching across devices despite different collecting views. The mapping relation acts as a mediator that reconciles the view differences and allows accurate tracking.
2Ease of manufacture
If feature matching association is performed without considering view differences, then the matching process is simple, but matching failure occurs when the same target object has different appearance features in overlapping regions
Solution Approach 1:
The patent performs preliminary coordinate transformation using the mapping relation before conducting feature matching. By pre-transforming the appearance features into a unified coordinate system, the system prepares the data in advance to ensure consistency during matching. This preliminary action prevents matching failures without complicating the subsequent matching process.
3Productivity
If tracking identifiers are assigned independently to each image, then the system can track objects in each view, but the same target object receives different tracking identifiers in overlapping regions
Solution Approach 1:
The patent merges the tracking identifier spaces from different image-collecting devices by establishing a unified tracking identifier system through the mapping relation. When objects are detected in overlapping regions, the system combines the tracking information from multiple devices and assigns consistent tracking identifiers, ensuring that the same target object receives the same identifier across all views while maintaining comprehensive tracking coverage.
Data Source
AI summary
The present disclosure provides a target tracking method and apparatus, relating to the field of image processing. According to embodiments of the present disclosure, a first image and a second image including a partially overlapping region synchronously collected by a first image-collecting device and a second image-collecting device are acquired, and first tracking detection blocks and second tracking detection blocks, for target tracking, of the first image and the second image are acquired respectively; the second tracking detection blocks are mapped to the first image according to a mapping relation between the first image-collecting device and the second image-collecting device, to obtain corresponding mapping blocks; and target objects in the overlapping region are fused according to intersection over union (IOU) and an appearance feature similarity level between the first tracking detection blocks and the mapping blocks.