Primitive Segmentation of Cyclic Patterns in Fabric Printing Based on Adaptive Template Matching

Abstract

The invention relates to a segmentation method of fabric printing cycle pattern primitives based on adaptive template matching, comprising the following steps: 1. the adaptive template selection: automatically determining an adaptive size template image in an image containing the printing cycle pattern; 2. template matching: template pattern in the step 1 is the most template, template matching iscarried out with the original image, and regions similar to the template pattern in the original image are found; 3, determining the size of the cycle: determining the size of the pattern cycle according to the positional relationship between the regions similar to the template pattern found in the step 2; 4. cyclic pattern primitive segmentation: according to the cyclic size determined in step (3), the original image is traversed and a complete pattern primitive conforming to human visual perception is segmented. The invention can automatically segment a complete pattern primitive from an image with a circulating pattern, avoids the shortcomings of slow manual segmentation speed, low efficiency and poor segmentation precision, and accelerates the flow of printing pattern development anddesign.

Description

【Technical field】 [0001] The invention relates to a method for segmenting fabric printing cycle pattern primitives based on adaptive template matching, so that a complete pattern can be automatically segmented from the printing fabric cycle pattern, and belongs to the technical field of digital image processing. 【Background technique】 [0002] Repetition is an important feature of printed fabric patterns, and most of the printed patterns on the market now have this feature. Now when most printing factories make printing patterns, they first scan the sample fabrics with cyclic patterns brought by customers into digital images; then use drawing software to manually intercept a minimum pattern cyclic base Then the designer draws the intercepted pattern cyclic primitives into a high-definition pattern suitable for plate making in the drawing software; finally, according to the customer's requirements, the drawn high-definition cyclic pattern primitives are combined into the cust...

AI Technical Summary

Problems solved by technology

However, most of the known cyclic pattern detection algorithms can only detect the cyclic size of the pattern in the horizontal and vertical directions, but cannot segment a complete cyclic primitive pattern that conforms to human visual perception from the cyclic pattern.

Also, these algorithms do not perform satisfactorily for cyclic patterns where rotation occurs in real production

Images