iDeBlurInfo: An Assessment and Evaluation Study of Image Deblurring Techniques

Quality of images used in a variety of systems contributes to the success of discovering beneficial details used in different fields of computing. The quality of the image plays a vital role in increasing the success rate of image recognition and related fields. Image quality can be degraded due to distortion, imperfection during image capture, inappropriate light balancing, and exposure to name a few. Various deblurring techniques were introduced to address blurred issues, transforming an image to lesser imperfections, successfully reading and/or identifying available details on it. This study aims to identify, define, and evaluate deblurring techniques in a way that demonstrates comprehension of the tradeoffs involved in the design. Significantly, the study compares various techniques that can be used to recover hidden details in the presence of image noise. The MA TLAB was used as a tool for testing and evaluating the deblurring techniques. PSNR, MSE, and computational time were used to compare each identified deblurring techniques. Based on the comparative performance results of each technique, in terms of accuracy measure, Lucy-Richardson got the highest PSNR ratio of 25.8312 that leads the best result while wiener filter got the lowest PSNR ratio of 21.3012. While, based on the MSE results, Wiener Filter got a higher rate of 0.0074; while blind deconvolution and Lucy-Richardson got the lowest value of 0.0026. And in terms of execution time, Wiener Filter performs the fastest computational time.