Design And Implementation Of Military Uniform Recognition System In Image

In recent years, domestic and foreign hostile forces have stepped up spying andpermeating our naval bases, military airport, missile position and other important militarytargets, some separatists broke into the barracks, more and more sentinels were killed byviolent terrorist, the situation at the border is still complex, and the Internet has become a newbattlefield. In this context, developing a kind of special security system to strengthen thedefense ability of the military administrative zone has become an urgent need. In view of thissituation, this paper tries to design a set of primary Military Uniform Recognition System(MURS). The MURS can not only be used for analyzing the important military target monitorscreen to identify the friend or foe in order to improve the performance of early warningsystem, but can also be used for searching sensitive military ID image on the web. The MURScan overcome low efficiency and other shortcomings of reading massive images by artificial.Therefore, this paper’s research possesses great innovation and pertinence, and has significantpractical value.Military uniform is a symbol of the Country, and the cloth and logo of the militaryuniform have distinctive features and strict rules to wear, so it is feasible to recognize themilitary uniforms in images. There are many kinds of military uniforms and the imagingpoints are different, so it is difficult to recognize the military uniforms. By using imageprocessing and recognition technology, this paper design and implement a MURS that candetect human faces in the image and analyze whether the clothing is a kind of the militaryuniforms, and then identify the types of the uniforms and rank level.The method of face detection is: segmenting skin regions by using two kinds of skincolor models, then screening the face region with shape feature and template matchingmethod, detecting and locating the eyes and lips area in the screened face region finally. Thepaper proposes to adjust the parameters of the color models in order to improve the accuracyof the face detection algorithm.The method of recognizing military uniforms is: firstly, detecting and positioning faces,secondly, segmenting reasonable shoulder and neck region, extracting the Tamura texturefeatures, blob feature, corner features and color features in order to judge the classification of the clothes, including whether they are military uniforms or not and which classification theyare, thirdly, dividing logo apparel by using different colors, shapes, spatial location and othercharacteristics according to the military uniforms classification, and verifying uniforms bymatching collar flowers templates, badges templates and other templates, and identifying theservices and the rank level of military uniform at the same time.Experimental results show good performance of the MURS designed in this paper, whichproved the correctness of the design of the overall framework of the system and the validity ofthe chosen features.