Object Tracking Apparatus Using Detection Difficulty for Frame Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object tracking technologies face challenges in maintaining high accuracy when multiple objects are tracked in complex scenarios, particularly due to decreased detection accuracy caused by similar features between objects and their surroundings, leading to reduced tracking precision.
Innovation Solution
An object tracking apparatus that utilizes priority camera information to select frames with lower detection difficulty degrees, predicting object positions in next frames, and calculating backside influence degrees to improve detection accuracy by avoiding frames with high overlap similarity, thereby enhancing tracking precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If object detection is performed from multiple camera views to track three-dimensional position, then tracking coverage is improved, but detection accuracy deteriorates due to occlusion and background similarity
Solution Approach 1:
The system performs preliminary action by predicting object positions in next frames before actual detection occurs. This allows the system to pre-identify regions of interest and prepare detection parameters, enabling more accurate detection when objects are visible and avoiding wasted processing when occluded.
Solution Approach 2:
The system applies dynamics by continuously updating detection difficulty degrees based on predicted positions and dynamically selecting frames for detection. The detection strategy adapts in real-time to changing scene conditions, switching between different camera views based on which provides the lowest detection difficulty at each moment.
2Duration of action of stationary object
If detection is performed in all available frames to ensure continuous tracking, then tracking continuity is improved, but detection precision deteriorates due to high detection difficulty in frames with occlusion or similar backgrounds
Solution Approach 1:
The system implements feedback by calculating detection difficulty degrees based on predicted positions and using this information to select optimal frames for detection. The detection results from selected frames are then used to update tracking state, which feeds back into position predictions for subsequent frame selection, creating a closed-loop system that continuously improves accuracy.
Solution Approach 2:
The system changes parameters by dynamically adjusting detection difficulty degrees based on predicted object positions, camera viewing angles, and background similarity. This allows the system to identify frames with favorable detection conditions (low detection difficulty) and prioritize those for actual detection, maintaining precision while ensuring continuity.
3Productivity
If frames with high overlap similarity are used for detection, then detection speed is improved, but detection accuracy deteriorates due to difficulty in distinguishing objects from background
Solution Approach 1:
The system performs preliminary action by predicting object positions and calculating detection difficulty degrees before actual detection. This pre-analysis identifies frames where objects are likely to be clearly visible with low background interference, allowing the system to select only those frames for detection, thereby maintaining both speed and accuracy.
Data Source
AI summary
An object tracking apparatus is configured to execute a tracking process, a prediction process, an influence-degree obtaining process, and a difficulty-degree obtaining process, wherein the influence-degree obtaining process is configured to obtain a backside influence degree representing that a detection of an object to be tracked is affected by other object that overlaps the object, wherein the difficulty-degree obtaining process is configured to calculate, for each object to be tracked, a detection difficulty degree for detecting the object from each of next frames captured by respective cameras, based on the backside influence degree, wherein the tracking process is configured to select the next frame that is included in a set of next frames in a pieces of video and from which the object is to be detected, based on the detection difficulty degree, and detect the object from the selected next frames.