Gas Leakage Infrared Imaging Enhancement Method Research

Some industrial leakage gas has intense absorbtion in medium or long-wave infrared region. Base on this mechanism, infrared imaging system, which has special sensitivity for weak temperature difference in atmospheric window, could detect gas escaped from leak points(leakage gas refers to industrial gas in this paper). In the infrared image, gas area shows gray or black because of background radiation absorption(Infrared image with leakage gas is named as leakage gas infrared image below). This research has done a lot of work around the leakage infrared image, including dealing with problems like random noise, weak detail and loss information in gas area colorization, and finally displays the gas area more distinctly.In order to solve problems mentioned above, we proposed a novel adapted wavelet denoise algorithm compositing image quality evaluation, SA(simulated annealing) and wavelet analysis. To promote the local contrast, we achieved an infrerad imaging detail enhance program based on gradient enhance and gray compression. Last, a gas colorizing project is established using optic flow theory.In the adapted wavelet denoise algorithm, image quality evaluation serves as penalty function of SA, the suitable wavelet threshold value is received by repeatedly iteration in SA. After human subjective judgment and objective evaluation, this algotithm is proved effective. In the precess of the infrerad imaging detail enhance program, we obtain the weak detail from the 14 bits original image, and then make the original image close to target image by minimizing gradient difference. The result is compared with HE(histogram equalization), and possess better detail information and local contrast. In the gas colorizing project, the gas area is firstly extracted by difference and watershed algorithms, and the velocity is calculated by optic flow. Then after color mapping and weighted stack based on saturability, the colored gas infrared image with velocity information is acquired, which not only enhances the visual effect but also contains background information. This search has made a new attempt on the basis of formed leakage gas infrared detection theory.