Blind image super-resolution (BISR) aims to restore corresponding high-resolution (HR) images from low-resolution (LR) images with unknown degradation, becoming a research hotspot in the field of computer vision. Due to the ill-posed nature of this problem, the rational selection and utilization of image priors have become key to solving it. However, existing deep learning-based BISR algorithms only employ neural networks to learn the mapping from LR to HR images in an end-to-end manner, only allowing the network to implicitly learn image priors, resulting in somewhat blurry super-resolution results. A deep learning-based BISR algorithm guided by sparsity and self-similarity priors is proposed to address the above issues. Initially, for various LR image inputs, a dynamic linear kernel estimation module is employed to effectively estimate the corresponding blur kernels; Subsequently, a deep unfolding deconvolution filtering module based on the Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) is utilized to explicitly model the sparsity prior of signal, achieving deconvolution restoration of the degraded images; Finally, a dual-path multi-scale large receptive field restoration module leverages the self-similarity prior of images for super-resolution recovery. Experimental results on the public Gaussian8 and DIV2KRK datasets demonstrate that, compared to existing methods, the proposed algorithm not only achieves the highest restoration metrics but also delivers superior visual quality in the recovered images.