An Image Enhancement Method for Gas Leak Detection Based on Infrared Imaging *

Infrared imaging technology is a useful tool for detecting gas leaks due to its significant advantages in detection spacial range, efficiency, and visualization. However, the original infrared images of gas leak traces often suffer from low contrast due to a small thermal radiation intensity difference between the gas plume and the scene. So it is difficult to directly separate them in the grayscale histogram distribution. Traditional enhancement methods lack targeted optimizations for traces of leaking gas. Therefore, enhancing the trace of gas leaks is crucial to the performance of gas leak detection systems based on infrared imaging. In this paper, we propose an image enhancement method based on guided filtering, the visibility restoration algorithm. The proposed method comprises three stages: image decomposition, base and detail layer enhancement, and image fusion output. We selected three infrared images of gas leakage taken in different scenes for experiments. There are obvious differences in the original images taken from the three scenes. We compare the image enhancement result of Histogram Equalization(HE), Contrast Limited Adaptive Histogram Equalization (CLAHE) and our proposed method. The image enhanced by our proposed has the highest values of quantitative indicators: average gradient(AG) and Enhancement Measure Evaluation(EME); which are three times higher than the original image. By visual observation, the gas leak plume in the images enhanced by our proposed method are obvious significantly.