Iterative Fault Correction for Object Tracking in Image Sequences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in accurately tracking objects over time, particularly when object characteristics change, leading to incomplete tracks due to faults in association between features across images.
Innovation Solution
A method and apparatus that iteratively corrects faults in track data by identifying candidate features and applying corrections to missing frames, using a combination of feature modules for identifying, merging, splitting, and clustering features, to improve the tracking of objects across a sequence of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If object tracking is performed using feature association across images, then tracking capability is achieved, but tracking accuracy deteriorates when object characteristics change over time
Solution Approach 1:
The system dynamically adjusts tracking parameters and re-evaluates feature associations at each frame based on current object characteristics. The tracking algorithm adapts to changing object properties by recalculating association metrics and updating track data in real-time, allowing accurate tracking despite temporal variations in object appearance.
Solution Approach 2:
The system changes tracking parameters such as association thresholds, feature weights, and confidence levels based on detected object characteristic changes. When objects undergo transformation, the system modifies these parameters to maintain optimal tracking accuracy, balancing the trade-off between maintaining track continuity and ensuring precise object identification.
2Duration of action of stationary object
If feature association is performed across all frames, then complete tracking is achieved, but computational complexity increases due to fault correction requirements
Solution Approach 1:
The system performs preliminary feature association in forward and backward passes through the image sequence before executing fault correction. By pre-establishing candidate associations and identifying potential faults in advance, the system reduces the computational burden of the correction phase, as it only needs to resolve specific identified issues rather than re-evaluate all associations.
Solution Approach 2:
The tracking process is segmented into distinct phases: forward pass for initial association, backward pass for validation, and iterative fault correction. This segmentation allows the system to handle different aspects of tracking separately, improving efficiency by focusing computational resources on correcting only the faulty associations rather than processing the entire track data repeatedly.
3Measurement precision
If iterative fault correction is applied to all tracks, then tracking accuracy improves, but processing time increases
Solution Approach 1:
The system applies fault correction iteratively but terminates the process early when a predetermined number of corrections have been applied or when corrections cease to produce significant improvements. This partial action approach achieves sufficient tracking accuracy without exhaustively correcting every potential fault, thereby reducing processing time while maintaining acceptable tracking precision.
Solution Approach 2:
The system automatically identifies and corrects faults in track data without requiring manual intervention or re-processing of entire sequences. The iterative correction mechanism self-adjusts by focusing on identified faults and stopping when convergence is achieved or maximum iterations are reached, optimizing the balance between accuracy improvement and time consumption.
Data Source
AI summary
There is provided a computer-implemented method and apparatus for determining temporal behaviour of an object. The method comprises receiving track data indicative of a plurality of tracks. Each track identifies an association between features corresponding to an object in each of a plurality of images forming image data representative of one or more objects. The method further comprises identifying, in at least some of the plurality of tracks, one or more faults, wherein each fault is associated with at least one respective fault image of the image data. The method further comprises iteratively performing a correction process on the track data. The correction process comprises selecting one of the plurality of tracks having a fault identified therein; determining one or more candidate features in each respective fault image, wherein each candidate feature is determined as a candidate for correcting at least one fault associated with one or more of the plurality of tracks; determining one or more candidate corrections for the selected track, wherein at least some of the candidate corrections are associated with one or more of the candidate features; selecting one of the candidate corrections for the selected track in dependence on a metric indicative of an effect of the candidate correction on the plurality of tracks; and applying the selected candidate correction to the selected track in dependence on one or more criteria.