Package design intelligent retrieval method and system based on big data
By dividing the packaging design drawing into non-overlapping blocks, using the maximum inter-class variance method and gradient intensity subdividing pixel points, the high and low quantization values are calculated, and the problem of inaccurate image feature expression in the prior art is solved, and the retrieval efficiency and accuracy are improved.
Patent Information
- Application Number
- CN202511072150.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-01
- Publication Date
- 2025-09-02
- Estimated Expiration
- 2045-08-01
AI Technical Summary
The existing block truncation algorithm based on absolute distance lacks the adaptive adjustment ability to the local structural characteristics of the image, resulting in insufficient accurate image feature expression and decreased retrieval accuracy.
By dividing the packaging design drawing into multiple non-overlapping blocks, the maximum inter-class variance method is used to dynamically obtain the optimal binarization threshold, combining the gradient intensity to subdivide the pixel points into three categories: strong boundary, weak boundary and non-boundary, and calculate the high and low quantization values based on the grayscale value and weight of the pixel points to form a double quantization feature representation.
It significantly improves the discrimination and matching accuracy of packaging design drawings, and improves the retrieval efficiency and accuracy in large-scale image databases.