Multi-Angle Image Recognition for Commodity Settlement Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-service commodity settlement methods based on image recognition technology suffer from reduced accuracy due to partial or complete obstructions of commodities, leading to missed detections and inaccurate settlements.
Innovation Solution
A commodity settlement processing method that involves capturing images of a settlement area from multiple angles, recognizing commodities in each image, and determining a final recognition result by comprehensively considering preliminary recognition results from different angles to improve detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single-angle image recognition method is used, then the device complexity is low, but the commodity recognition accuracy deteriorates due to obstructions
Solution Approach 1:
The patent transitions from single-angle (2D) image capture to multi-angle (3D spatial) image capture. By arranging multiple cameras at different positions around the settlement area, the system captures commodity images from multiple dimensional perspectives, enabling comprehensive recognition even when commodities are partially obscured from any single viewpoint.
Solution Approach 2:
The patent divides the settlement area into multiple monitoring zones, each covered by a specific camera positioned at a determined location. This segmentation approach ensures that different cameras capture different portions or angles of the settlement area, and through data fusion of these segmented views, the system achieves complete and accurate commodity recognition.
2Measurement precision
If multiple cameras are deployed to capture images from different angles, then the commodity recognition accuracy is improved, but the device complexity and cost increase
Solution Approach 1:
The patent optimizes the parameters of the camera system by determining specific positions and shooting angles for each camera based on the settlement area layout. Rather than arbitrarily deploying cameras, the system calculates optimal parameters (position coordinates, viewing angles, focal lengths) to achieve comprehensive coverage with the minimum necessary number of cameras, balancing accuracy improvement with device complexity control.
3Measurement precision
If multiple shooting angles are used to capture commodity images, then the detection accuracy is improved, but the processing time and complexity increase
Solution Approach 1:
The patent performs preliminary actions by pre-determining the optimal positions and shooting angles of cameras before actual commodity settlement. The camera parameters and viewing angles are pre-calculated based on the settlement area geometry, so that during actual operation, images can be immediately captured from the predetermined optimal angles without requiring real-time computation or adjustment, thus reducing processing time.
Solution Approach 2:
The patent creates multiple copies of the settlement area view from different virtual perspectives through the multi-camera system. Each camera captures a copy of the settlement scene from its specific angle, and these copies are then fused to reconstruct the complete commodity information, enabling accurate detection without requiring complex real-time analysis of a single comprehensive view.
Data Source
AI summary
A commodity settlement processing method is provided. The method includes obtaining N first images by shooting a settlement area from N shooting angles, the N is a positive integer that is greater than 1. Commodities in each of the N first images are recognized and a preliminary recognition result corresponding to each first image is obtained. Once a final recognition result is determined based on N preliminary recognition results that have been obtained, the commodities in the settlement area are settled according to the final recognition result.