This paper addresses the single image scale-up problem using sparse-representation modeling. The goal is to recover an original high-resolution image from a blurred, down-scaled, and noisy low-resolution version. The problem is highly ill-posed, necessitating a prior for regularization. The authors build on a successful algorithm by Yang et al., assuming a local *Sparse-Land* model on image patches for regularization. They introduce several modifications to simplify the process, including a different training approach for the dictionary pair and the ability to operate without a training set by bootstrapping from the given low-resolution image. The results are demonstrated on real images, showing both visual and PSNR improvements. The paper also discusses the role of blur in the scale-up process and compares it with other approaches that separate deblurring and upsampling tasks.This paper addresses the single image scale-up problem using sparse-representation modeling. The goal is to recover an original high-resolution image from a blurred, down-scaled, and noisy low-resolution version. The problem is highly ill-posed, necessitating a prior for regularization. The authors build on a successful algorithm by Yang et al., assuming a local *Sparse-Land* model on image patches for regularization. They introduce several modifications to simplify the process, including a different training approach for the dictionary pair and the ability to operate without a training set by bootstrapping from the given low-resolution image. The results are demonstrated on real images, showing both visual and PSNR improvements. The paper also discusses the role of blur in the scale-up process and compares it with other approaches that separate deblurring and upsampling tasks.