Abstract:
The use of infrared cameras for gas visualization, also named gas cameras, has facilitated the search of escaped environmentally harmful gases in the petroleum and chemic...Show MoreMetadata
Abstract:
The use of infrared cameras for gas visualization, also named gas cameras, has facilitated the search of escaped environmentally harmful gases in the petroleum and chemical industry, since such cameras allow the detection of gases in a remote way. One way to locate a gas cloud spatially is by setting up two gas cameras in a stereo arrangement and fusing their information. Since gas clouds are semi-transparent non-stationary textures in gas camera images, current methods for depth measurements from stereo visual systems may not succeed or be inaccurate. Furthermore, spatial jitter in the images originated by the internal cooling engine of a gas camera and the radiometric and temporal inaccuracies of the system can affect the performance of the reconstruction method. In order to calculate accurate spatial information with a stereo gas camera system, in this work a characterization of the geometric, radiometric and temporal inaccuracies of the system and a calibration procedure were carried out. After evaluating the obtained results by means of a case study, it has been shown that the location and spatial information of gas clouds can be achieved automatically through a calibration and spatial jitter compensation in the stereo images.
Published in: 2018 IEEE Sensors Applications Symposium (SAS)
Date of Conference: 12-14 March 2018
Date Added to IEEE Xplore: 12 April 2018
ISBN Information: