Wavelet Threshold Denoising of Thermal Image from Transmission Joints

In the infrared diagnosis of power equipment, the thermal imagery collected from the transmission joint can be affected by several external interferences, which lead to bad contrast and mixed with various types of noise. This phenomena is not conducive to make fault diagnosis accurately, therefore the denoising and enhancement in thr thermal imagery are very important. Several external factors which affect thermal image quality are summarized firstly, and then the advantage and disadvantage of the original methods of the enhancement and denoising in the thermal imagery are analyzed. What´s more, we also present a new method based on local histogram equalization then denoised through the wavelet threshold. Finally, the presented method is compared and analyzed with the enhancement in spatial domain and wavelet threshold denoising. Results show that the presented method can enhance the contrast and remove the thermal imagery noise at the same time more effectively, thus help the infrared diagnosis of power equipment.