In the process of automatic strip steel surface defect detection, due to the influence of the camera itself and the camera environment, the acquired images will have uneven brightness, which has a greater impact on the subsequent defect image segmentation and recognition. Based on this, this paper proposes a brightness unevenness correction method based on improved homomorphic filtering, which combines the characteristics of strip surface defects to reselect the cutoff frequency of the filter, and uses different filter functions for the high and low frequency components of the strip image. After processing, the processed high and low frequency components are used to reconstruct the brightness-corrected image. The experimental results show that the improved homomorphic filtering in this paper has higher image quality after brightness correction. The image quality parameter Q value after brightness correction is used to evaluate it. The average Q value of the method in this paper reaches 0.9014, which is higher than the parameter values of other methods. high.