Leaked gas cloud cluster detection method based on thermal imaging video

Abstract

The invention discloses a leaked gas cloud cluster detection method based on a thermal imaging video, and belongs to the field of gas detection. The leakage gas detection model based on the shape features of the gas infrared image is obtained by using the support vector machine and SIFT feature training, automatic detection of the leakage gas is realized, the leakage condition of the gas is displayed in a video form by using the visibility of the leakage gas in an infrared scene, and the method has the advantages of large-range imaging and direct positioning of a leakage source. The SIFT features are used as feature data of leaked gas to be input into the support vector machine, the advantages of small feature data size, low model complexity and short training and detection time of the SIFT features are fully utilized, and the detection accuracy of the leaked gas is improved by reducing the complexity of a leaked gas detection support vector machine model. And the time cost and the storage space cost of the leakage gas cloud cluster detection method based on the thermal imaging video are compressed. The method has the advantages of being low in cost and good in real-time performance while meeting the requirement for automatic detection of leaked gas.

Description

technical field [0001] The invention relates to a method for detecting leaked gas clouds based on thermal imaging video, belonging to the field of gas detection. Background technique [0002] As a high-quality, clean fuel and chemical raw material, natural gas has become an important part of my country's energy structure. However, major natural gas leakage accidents occur every year in various places, which makes the technical research and instrument development related to leaked gas detection imminent. Although the traditional contact gas leak point detector has high sensitivity, its detection efficiency is low. In recent years, the gas leakage detection technology based on infrared imaging system has developed rapidly. Compared with the traditional detection method, this method images the environment and can directly observe the location and diffusion of gas leakage from the screen, which not only improves the detection efficiency but also realizes long-term monitoring an...

AI Technical Summary

Problems solved by technology

The existing research on thermal imaging detection algorithms for leaked gas mainly refers to the target detection model of visible light scenes, using convolutional neural network (CNN) training to directly train the infrared image of leaked gas, but lack of research on the characteristics of infrared image of leaked gas, resulting in model training and Leakage gas detection requires a large amount of calculation and takes a long time, and the resulting convolutional neural network model for gas leakage detection has high complexity, which is difficult to meet the needs of real-time detection of gas leakage

Images