Research On Mosaic Method Of Digital Camouflage

As a new means of camouflage, digital camouflage has refined simulated background which has advantages that include against the detection of aerial shot and remote satellite sensing.However, when painting camouflage on ground martial targets, there usually is a seam at stitches of the patterns thus exposing camouflaged targets and reducing the effect of camouflage. Therefore,resolving the stitching issue of patterns is of great significance to improving the camouflage effect.Given the distinctiveness of digital camouflage patterns, namely fewer color choices and fragmented patterns, common stitching technology based on SIFT and SURF is no longer applicable. In light of this issue,the paper proposed a new stitching method for digital camouflage patterns. Basically, it first converts camouflage patterns into gray scale images and then divides and labels the image with minimal camouflage unit; then,identifies a camouflage unit module to acquire corresponding index of each gray scale label after module calculation; afterwards, assigns the value of the largest gray scale label to the camouflage plate at the stitching seam; at last,through the gray scale value of the camouflage pattern,gets the corresponding main camouflage color. Following the above steps, one row or several rows of transition patterns will be acquired to add or replace camouflage patterns at the stitching seam.This paper adopts two approaches to test the camouflage effect of the mosaic digital camouflage patterns. One is the single Peak Signal-to-Noise Ratio index and the other is the multiple standards for similarity. Experiment indicates that the algorithm proposed in this paper will effectively reduce the exposure rate of camouflage so as to achieve the natural transition from the original camouflage pattern.