Weighted variational regularization image dehazing algorithm based on non-local prior

In hazy weather conditions, the quality of images taken outdoors will be degraded, so researching image dehazing has great practical significance. Therefore, an image dehazing algorithm based on non-local prior weighted variational regularization is proposed. Since the performance of the image dehazing algorithm depends largely on the accuracy of the estimated transmission, a classical non-local prior algorithm is first adopted to make a rough initial estimate of the transmission. Then a weighted variational regularization model is proposed to further optimize the initial transmission and solve the model by the alternating direction multiplier method. Finally, a haze-free image can be recovered directly from the atmospheric scattering model by using the optimized transmission. The experimental results verify that the proposed algorithm has higher image dehazing performance than several commonly used algorithms.