Blind Image Blur Type Estimation and Image Deconvolution Techniques

These days, one of the main challenges with photos is distortion. This has an impact on numerous fields, including microscopy, medical imaging, astronomy, remote sensing, and photography. Many factors, such as vibration from hand movements and satellite launches, noise in the image, unfavorable image/environment conditions, and fast object movement, can cause images to get obscured. It is necessary to develop a method that can address the issues raised above and enable actions to minimize image obscureness. Blur detection is the first step that must be completed for any blind image restoration out of multiple steps in the process. This paper presents a comparison of different methods that can be used to determine parameters of blur from a corrupted image by utilizing some features like Zernike Moment, Moment Invariants, and Histogram of Oriented Gradients. Additionally, this research compares various linear and nonlinear restoration methods. The Structural Similarity Index (SSIM), Peak Signal-to-Noise Ratio (PSNR), assessment of blur, and forms of blur were the basis for the analysis and comparison.