Label Verification via OCR and Alternating Barcode Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Decoding barcodes using mobile devices with cameras is computationally intense, leading to lag in web-based applications and inaccurate tracking due to motion blur and limited computational resources.
Innovation Solution
Implement a tracking scheme that decodes barcodes in alternating frames while tracking their positions in between, using a correlation filter to predict and update positions without continuous decoding, reducing computational load and improving tracking accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If barcode decoding is performed continuously in every frame, then decoding accuracy is improved, but computational load increases causing lag and energy consumption
Solution Approach 1:
The system performs barcode decoding only in alternating frames (every other frame) rather than continuously in every frame. This periodic decoding approach reduces computational load and processing requirements while maintaining effective tracking capability, as the correlation filter can accurately predict barcode positions in between decoded frames
Solution Approach 2:
The correlation filter is pre-trained with barcode features before tracking begins. This preliminary preparation allows the filter to quickly predict barcode positions in subsequent frames without requiring full decoding computation, enabling efficient real-time tracking with reduced processing demands
2Reliability
If barcode decoding is performed in every frame, then tracking accuracy is improved, but energy consumption increases
Solution Approach 1:
The system adopts periodic decoding at alternating frames rather than continuous decoding in every frame. This reduces energy consumption significantly while maintaining reliable tracking through the correlation filter's ability to predict positions between decoded frames based on learned barcode characteristics
Solution Approach 2:
Instead of performing full barcode decoding in every frame, the system creates predicted copies of barcode positions using the correlation filter. These predicted positions are sufficient for tracking purposes and avoid the high energy cost of repeated full decoding operations
3Measurement precision
If full decoding is performed continuously, then decoding precision is maintained, but device complexity and processing requirements increase
Solution Approach 1:
The tracking system is divided into two distinct components: (1) periodic full decoding performed only in alternating frames to maintain precision, and (2) continuous lightweight position prediction using the correlation filter in between decoded frames. This segmentation reduces overall processing complexity while preserving decoding accuracy where needed
Solution Approach 2:
The correlation filter serves as an intermediary between full decoding operations. It takes decoded barcode features from selected frames and generates predicted positions for intermediate frames, acting as a computational bridge that reduces the need for frequent full decoding while maintaining tracking continuity and precision
Data Source
AI summary
Image analysis is used to provide label verification. An image of a label within a scene is acquired. The label includes a barcode and characters. The barcode on the label is decoded to receive an identity of an object. Optical character recognition is performed to recognize the characters on the label. Recognized characters are compared to information about decoded barcode to verify the label is correct.