A Multi-Scale Fusion Bogie Image Enhancement Algorithm Based on Collaborative Optimization of Image Features

The structural integrity and performance stability of the bogie directly affect the safety of train travel. Currently, bogie safety monitoring based on computer vision is increasingly favored by users.But the problems of low image contrast, noise interference and blurring caused by uneven lighting and occlusion in the bogie detection environment need to be urgently addressed.This study proposes A multi-scale fusion bogie (MFB) image enhancement algorithm based on collaborative optimization of image features. The algorithm first uses the Retinex model for illumination estimation, combined with gamma correction to achieve brightness equalization, in order to improve the overall contrast of the image. Subsequently, the color distribution of the image is optimized through color gain weighting and white balance correction, enhancing the naturalness and fidelity of colors. Further use a sharpening method based on Laplacian operator to enhance edge features, and layer the image through multi-scale dehazing and denoising modules to reconstruct the image layer by layer. Finally, through a multi-scale weighted feature fusion strategy, images with color enhanced, edge sharpened, and noise reduced are weighted and fused to generate high-quality enhanced images. The experimental results show that compared with traditional image enhancement methods and deep learning based enhancement methods, our algorithm exhibits significant advantages in both subjective visual effects and objective indicators, especially in detail restoration and color fidelity.