Fractional-order Retinex-based low-light image enhancement fusion algorithm for energy meters

For the haze, rainy days and night and other detection environment, the smart meter collection image parameters difficult to identify the problem, A fractional-order improved Retinex-based low-light image enhancement fusion algorithm is proposed; The image is enhanced by the fractional-order improved Retniex-based low-light enhancement algorithm, and the environmental noise is filtered out by the Fast ABF algorithm, and the second enhancement process is performed; the original image, the first low-light enhanced image, the Fast ABF enhanced image and the second low-light enhanced image are decomposed by the Laplace pyramid and multiplied by the corresponding Gaussian weight matrix, and the final enhancement is obtained by fusion image; Among them, the self-built smart meter dataset is used, and the YOLOv5 model is adopted to extract the target meter dial area; the results show that the algorithm outperforms other comparison algorithms in terms of image naturalness retention and image sharpness, so that the detection accuracy is improved.