Retail Object Recognition via Dual-Resolution Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Object recognition in retail environments is challenging due to high variability in product attributes and frequent introductions of new products, which current systems struggle to adapt to effectively.
Innovation Solution
An image processing system that includes a database of visual identifiers and a server-based system that receives images from imaging devices, attempts to recognize products, and requests additional high-resolution images when necessary to update the database and improve recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If current object recognition systems are used in retail environments, then they can identify products, but they fail to adapt to high variability in product attributes and new products effectively
Solution Approach 1:
The system performs preliminary actions by capturing multiple images at different resolutions before recognition is attempted. Low-resolution images are processed first for quick matching, and only when needed are high-resolution images captured and processed to update the visual database, enabling the system to adapt to new products proactively
Solution Approach 2:
The system dynamically adjusts its operation mode based on recognition confidence levels. When product variability is detected or recognition fails, the system transitions from using only low-resolution images to capturing and processing high-resolution images, allowing flexible adaptation to changing retail environment conditions
2Measurement precision
If high-resolution images are always captured and processed, then recognition accuracy improves, but system complexity and processing time increase
Solution Approach 1:
The image processing is segmented into two stages: first, low-resolution images are processed for rapid visual identifier matching; second, high-resolution images are processed only when needed to update the visual database. This segmentation reduces overall system complexity by avoiding constant high-resolution processing
Solution Approach 2:
The system changes the resolution parameter dynamically based on recognition needs. Low-resolution images (e.g., 320x240 pixels) are used for initial matching to reduce computational load, while high-resolution images are used selectively for database updates, optimizing the balance between accuracy and complexity
3Productivity
If low-resolution images are used for recognition, then processing speed increases, but recognition accuracy decreases
Solution Approach 1:
Low-resolution images are processed as a preliminary step to achieve quick recognition matches. The system attempts identification using these fast-to-process images first, and only proceeds to high-resolution processing when the preliminary attempt fails, thus maintaining both speed and accuracy
Data Source
AI summary
The present disclosure provides a method of image processing comprising: obtaining by an imaging device a low resolution version and a high resolution version of a retail image, the high resolution version of the retail image being a temporary file to be erased automatically after a predetermined time period; transmitting to a server the low resolution version of the retail image; upon receipt of a request from the server, the request including data representative of a contour of an unidentified item in the low resolution version of the retail image, cropping a high resolution item image from the high resolution version of the retail image, the high resolution item image corresponding to the contour of the unidentified item; and transmitting the high resolution item image to the server thereby enabling updating an item database.