Evaluation of Image Deblurring Techniques

Abstract

Degradation of images is one of the major problems in image processing. Blur in images is an unwanted reduction in bandwidth which degrades the image quality and it is difficult to avoid. Blur occur due to atmospheric turbulence as well as improper setting of camera. Along with blur effects, noise also corrupts the captured image. Restoration of image is a technique to get rid of the blur from the degraded image and recover the original image. Blur can be of various types like Gaussian blur, motion blur etc. Now a day's there are various different techniques and methods have been proposed to deblur a degraded image. For specific types of blur there are specific methods to remove it. Image restoration has applications in various different-different fields like medical imaging, forensic science, and astronomy. In this paper, we will discuss various image deblurring techniques and their analysis of performance.

Simulating Study on Observation of Space Point Target from a Shipborne Optoelectronic System

“…The main reason is that after 4 iterations the motion-blur is compensated effectively, but the noise is not eliminated or weakened. Furthermore, the image noise is enhanced as the iteration number increases [8]. So the effect of deblurring is degraded when the iteration number is beyond 4.…”

“…But the noise degrades the definition of the image badly, and RL algorithm does not remove the noise at all. On the contrary, RL algorithm enhances the image noise as the iterative number increases [8]. So the influence of noise on recovering of motion-blurred image should be investigated.…”

Palm Vein Technology for Biometrics

“…60 different angles are computed in Radon space to enhance low frequency components. Discrete Cosine Transform (DCT) is then applied to obtain frequency features in Radon space [13]. 25% of DCT coefficients are summed to form feature vector.…”

Thermal Image Reconstruction Using Deep Learning

“…Another limitation of using a thermal camera is image blurring, which is a common image degradation problem that occurs in real scenarios for various reasons, such as defocus, unbalance, camera shaking, motion, and noise. These image degradations are usually solved using a nonblind deconvolution method, such as Wiener filter [2], Lucy-Richardson algorithm [3], and regularized filter [4], and a blind deconvolution method [5]. However, blurs in images obtained with a common thermal camera are more serious than those in images obtained with a common visible-light camera.…”