Temperature Recognition System Of Thermal-paint Color Images Based On Color Image Processing Technology

Temperature measurement of high-temperature objects has been widely applied in many fields such as aeronautics, astronautics, military and industrial production. It is also an important research subject both in theoretic and industry practitioners. It will be very significant and meaningful for theory and practices to develop a temperature recognition system of thermal-paint images by using the image processing technology.For the distribution of the noise in thermal-paint images, a square neighborhood peer group filtering algorithm is used to filter the noise after comparing with traditional image filtering methods. The proposed algorithm can not only remove the noise but also remain the edges effectively so as to get the better filtering results.With comparison and analysis of the existing color image quantization, LUV color space which accords with visual consistency is adopted to develop three kinds of segmentation algorithms by using color information, spatial information and the distribution of color densities. The three algorithms are fast image segmentation algorithm based on the V-S algorithm, improved image segmentation algorithms based on K-means method and image segmentation based on mean shift method. The algorithms have effectively achieved image quantization of thermal-paint images. Finally, a direct identification method and a curve identification method based on the features of color temperature are used to measure the temperature of paint images.The thermal-paint recognition system was developed, which was designed to achieve smooth image filtering, quantitative segmentation, temperature identification and the data management of thermal-paint sample data. The experiment with the thermal-paint recognition system shows that thermal-paint images can be effectively processed and analyzed and the temperature can be recognized from the thermal-paint images. Deviation of the temperature recognition is controlled in ten centigrade degree, which is accorded with industrial performance index.