The invention provides a laser spot image denoising method based on a deep learning convolutional neural network. The laser spot image denoising method comprises the following steps: S1, image acquisition: acquiring an original spot grayscale image through an image acquisition device; S2, light spot image noise classification processing: inputting the original light spot grayscale image into a classifier network to perform label prediction to obtain a label image, and performing noise classification to obtain a sparse clear image; and S3, light spot image denoising processing: inputting the sparse clear image and the original light spot grayscale image into a regression network, and reconstructing a light spot image to obtain a denoised light spot image. The laser spot image denoising method performs noise elimination on the light spot image by using the regression network, and feeds the sparse clear image and the original light spot image to the regression network together to reconstruct a denoised image so that the accuracy of eliminating random noise of the light spot image can be enhanced.