Infrared Image Enhancement for Photovoltaic Panels Based on Improved Homomorphic Filtering and CLAHE

To address the problems of low contrast and low illumination of infrared images of photovoltaic panels, an infrared image enhancement for photovoltaic panels is proposed. Firstly, in order to improve the overall brightness and contrast of the infrared image, a homomorphic filtering algorithm based on improved transfer function is designed, which constructs a transfer function with a similar structure to the homomorphic filtering profile. Secondly, using the contrast limited adaptive histogram equalization (CLAHE) algorithm fused with gamma correction to further process the image, which not only overcomes the defects of weak details and uneven brightness of the image enhanced by homomorphic filtering, but also improves the clarity and anti-interference of the image. The experimental results show that the proposed algorithm can effectively enhance the visual effect of infrared images, and then improve the integrity of photovoltaic panels in manually labeled images and the detection accuracy of photovoltaic panels.