三种色彩模式下火灾图像偏色度量化方法研究  

Research on Partial Chromaticity Quantization Method of Fire Image in Three Color Modes

作　　者：李海 孙鹏[2] 熊升华 伍毅[1] LI Hai;SUN Peng;XIONG Sheng-hua;WU Yi(Civil Aviation Flight University of China,Deyang Sichuan618307,China;Criminal Investigation Police University of China,Shenyang Liaoningl 10035,China)

机构地区：[1]中国民用航空飞行学院,四川德阳618307 [2]中国刑事警察学院声像资料检验技术系,辽宁沈阳110035

出　　处：《计算机仿真》2021年第8期470-477,共8页Computer Simulation

基　　金：国家重点研发计划资金(2018YFC0810600);国家重点研发计划专项(2017YFC0822204);中央高校基本科研业务费重大项目(D2018005);中国民用航空飞行学院青年基金项目(Q2020-182)。

摘　　要：环境的复杂色彩严重影响了火灾现场救援过程中目标识别的准确性与及时性。为此提出一种面向三种色彩模式的基于欧式距离的火灾图像偏色度量化方法。首先通过三种情景模拟火灾现场,利用相机拍摄不同时空下的火灾现场场景,同时拍摄不同时空常规现场场景,形成火灾现场偏色检测图像库;其次面向RGB、HSV、YCbCr三种色彩模式提出欧式距离法火灾图像偏色度量化方法,同时针对火灾图像基于图像库进行偏色度检测;最后,利用MATLAB软件进行统计分析,验证方法的可靠性。仿真结果表明,三种色彩模式下的基于欧氏距离的火灾图像偏色度量化方法能够有效检测各类情景图像,检测效果较好,能够避免单一色彩模式下火灾图像偏色度量化带来的局限性。The complex color of the environment seriously affects the accuracy and timeliness of target recognition during the fire rescue process.To this end,a partial chromatiocity quantization method based on Euclidean distance for the three-color mode is proposed.Firstly,the fire scene was simulated by three scenarios,and the fire scene in different time and space was taken by the camera.At the same time,the scenes in different time and space were photographed to form the image field of the color field detection of the fire scene.Secondly,the three color modes of RGB,HSV and YCbCr were proposed.Distance method was used to measure the color of the fire image,and the color image was detected based on the image library.Finally,the statistical analysis was performed by MATLAB software to verify the reliability of the method.The simulation experiments show that the Euclidean distance-based fire image color metric quantization method in three color modes can effectively detect various scene images,and the detection effect is good,which can avoid the measurement of the fire image color cast in single color mode.Limitation.

关 键 词：火灾图像 偏色因子 偏色度量化 欧式距离 色彩模式 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]