CH4 gas leakage detection method for low contrast infrared images

The leakage of CH4 gas not only pollutes the environment, but also poses a serious threat to social safety and human health. Infrared imaging is an effective method for detecting CH4 gas leakage. It has the advantages of convenient operation and noncontact. However, compared with general infrared images, gas infrared images have lower contrast and more blurred edges. When the CH4 leakage is small, the detection methods are easily disturbed by noise. To solve this problem, this paper proposes a detection method for CH4 gas leakage based on a Gaussian mixture model and a Kernel density estimation model. First, in the preprocessing part, the improved adaptive anisotropic diffusion filter is used to denoise and preserve the edge of the image. Then, we employ the fast and robust fuzzy c-means clustering (FRFCM) guided contrast limited adaptive histogram equalization (CLAHE) method to enhance the details of the infrared image to improve the local contrast. Finally, the Gaussian mixture model and improved Kernel density estimation model are cascaded to segment the gas area and mark leakage areas. The experiment results demonstrate that our proposed method has more significant segmentation and anti-jamming ability than other algorithms under indoor conditions of 30 mL/min and 0.8 m. It can effectively detect the CH4 leakage area and has a higher accuracy rate and lower false positive rate.