Lightweight Image Super-Resolution Reconstruction Based on Edge Detection

Image super resolution reconstruction is to generate high resolution image from low resolution image, which plays an important role in many fields. In this paper, a lightweight image Super-Resolution reconstruction method based on edge detection is proposed, which is improved by the classical Convolutional Neural Network (SRCNN) model. By optimizing the convolution kernel structure of SRCNN and introducing Sobel and Laplace operators to detect and enhance the edge of the image, the edge details and the overall visual quality of the reconstructed image are improved. The experimental results show that the improved model can significantly improve the performance of image super-resolution reconstruction while maintaining low computational complexity. PSNR is improved by more than 10% in both x2 and x4 scales, especially in edge sharpness and detail restoration.