Multi-Hierarchy Video Item Tracking Using Dynamic Time Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current item detection schemes based on video data face challenges in accuracy, often misidentifying items when they are moved briefly, leading to repeated identification and low robustness.
Innovation Solution
An apparatus and method that utilize a detector, tracker, and classifier to perform multi-hierarchy decision using different time windows for item identification in video data, ensuring accurate tracking and classification by preprocessing and post-processing image frames to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video data detection is used for item identification, then detection speed is improved and barcode fraud is prevented, but detection accuracy deteriorates when items are moved briefly
Solution Approach 1:
The system dynamically adjusts the time window size based on the detection scenario and item characteristics. For items that may be temporarily moved, a larger time window is used to track the item across frames and maintain identification consistency, while for stable items, a smaller time window provides more responsive detection. This dynamic adaptation resolves the contradiction by making the detection system flexible enough to handle both high-speed detection and accuracy-critical scenarios.
Solution Approach 2:
The system performs preliminary tracking of detected items across multiple frames before final identification. By pre-processing the detection results through temporal tracking and using multi-hierarchy decision-making, the system prepares refined identification data before final classification, thereby improving accuracy without significantly impacting detection speed.
2Device complexity
If traditional single-hierarchy detection is used, then system complexity is low, but item identification accuracy deteriorates when items are temporarily moved
Solution Approach 1:
The detection system is segmented into multiple hierarchical levels, with each level performing specific detection tasks at different time window scales. This segmentation allows the system to handle temporary item movements by checking multiple hierarchy levels, improving identification accuracy while keeping each individual detection module relatively simple and manageable.
Solution Approach 2:
The multi-hierarchy decision structure implements a nested decision-making framework where smaller time window detections are nested within larger time window analyses. This nested approach allows the system to maintain detailed local detection accuracy while incorporating broader temporal context, resolving the contradiction between system complexity and identification accuracy.
Data Source
AI summary
The embodiments of the present disclosure provide an apparatus for identifying items, a method for identifying items and an electronic device. The apparatus includes: a detector configured to detect one or more items in a reference area in one or more image frames in video data; a tracker configured to track an item detected in multiple image frames, wherein multi-hierarchy decision is performed on the item in the multiple image frames by using different time windows; and a classifier configured to identify the item according to a decision result of the tracker. Thereby, even if an item is moved briefly in some scenarios, the item will not be identified as two different items, which can reduce a situation in which the item is identified repeatedly and improve accuracy and robustness of item detection.