Abstract：Automation and intelligence are the development trend of substation. More and more automatic inspection robots are used in inspection work. Infrared image leakage detection technique is widely used in detecting SF6 gas leakage in electrical equipment. It can be used in local leakage searching and large-scale patrolling. Therefore, it is necessary to study the automatic location algorithm of infrared leakage detection. First, gray processing is applied to infrared image and Gauss blur is used to eliminate the noise. Threshold transformation is used to recognize the device area in the first frame image and obtain mask. Then, accumulated interframe difference method is used to get the cumulative changes in infrared images. Finally, the interferences outside the device area are removed, and the leakage points are located. The experiment shows that the proposed method has fast computing speed and good anti-interference ability, which makes it suitable for real-time video processing. This method can be used in both automatic substation patrol robot and handheld infrared image leakage detector.
黄柏, 杨帆, 邓剑平, 毕凡. 基于累积帧间差分法和掩膜的SF6红外检漏视频定位算法研究[J]. 电气技术, 2022, 23(7): 104-108.
HUANG Bo, YANG Fan, DENG Jianping, BI Fan. Study of an accumulated interframe difference and mask based SF6 leakage infrared video location method. Electrical Engineering, 2022, 23(7): 104-108.
2022, Vol. 23 
