Volume 8, No. 5, May-June 2017
International Journal of Advanced Research in Computer Science
RESEARCH PAPER
Available Online at www.ijarcs.info
© 2015-19, IJARCS All Rights Reserved 2442
ISSN No. 0976-5697
Blind Image Deconvolution using Frequency Spectrum in Motion Blur Estimation
Aarti Chugh
Assistant Professor, Department of Computer Science,
Amity University Haryana, India
Charu Jain
Assistant Professor, Department of Computer Science,
Amity University Haryana, India
Abstract: Image deconvolution refers to the act of minimizing the blur quantity from a blurred image and discovers the sharp and clear original
image [7].The blind image deconvolution (BID), refers to the task of separating two convolved signals.We have taken motion blur estimation
method which will extract the features from the frequency domain of the image. This method has low complexity as compare to other estimation
methods like radon, steerable filter methods.
Keywords: Image Deconvolution, Point Spread Function (PSF), Frequency, Motion Blur
I. INTRODUCTION
Image deconvolution refers to the act of minimize the blur
quantity from a blurred image and discover the sharp and
clear original image. It uses a point spread function (PSF
also known as blur kernel) to deconvolve the blurred image.
There are four types of blur effects in digital camera
[14].Average blur is a tool to remove noise and specks in an
image. It is used when noise is present over the entire image.
Second is motion blur. It is degradation in a photograph of
any moving object or imaging system itself. Motion blur
causes significant degradation of the image[1][11]. A
gaussian blur is the result of blurring of an image by
gaussian function. This blur has good capability to reduce
image noise. Hence it iscommonlyused in graphics software.
The Gaussian blur effect is a filter that blends a specific
number of pixels incrementally, following a bell-shaped
curve [9]. Gaussian blur is also used as a pre-processing
stage in computer vision algorithms in order to enhance
image structures at different scales. Out-of-focus Blur
occurs when a camera images a 3-D scene onto a 2-D
imaging plane. This causes some parts of the scene in focus
while other not. There are many methods for deblurring a
blurred image.
The blind image deconvolution (BID), refers to the task of
separating two convolved signals, iand d, when both the
signals are either unknown or partially known[7]. A lot of
research has been done exploring the various methods for
image deconvolution as blind techniques. But still, it is a
critical and challenging problem for the researchers.
II. LITERATURE SURVEY
A.S. Mane, Prof. Mrs. M. M. Pawar [1] has worked on a
new effective blind deconvolution algorithm which can
effectively remove complex motion blurring from natural
images without requiring any prior information of the
motion-blur kernel. In degradation model for blurring
image, the image is blurred using filters and an additive
noise. After this, apply restoration process to restore image
using blind deconvolution technique. Estimate PSF of
different size of initially blurred image then apply canny
edge detector to remove ringing effects from restored image.
Limitation of this algorithm was large number of steps were
there of deconvolution process.
Ashwini M. Deshpande and Suprava Patnaik [2] has
proposed an algorithm based on sparse representation of
images and adaptive dictionary learning, features one of the
successful repositories in removal of uniform and non-
uniform motion blur from a single blurred-noisy image.
Comparative performance analysis performed with some of
the leading deblurring algorithms proves the effectiveness of
the proposed method in handling blurs like linear blur,
camera shake, hand shake etc. These algorithm works for
both uniform and non-uniform motion blur.
Sunghyun Cho, Yasuyuki Matsushita, Seungyong
Lee[3][12] has proposed a new method for removing non-
uniform motion blur from multiple blurry images. First, it
proposes a new approach to restoring images that are
contaminated by spatially-varying motion blurs. Second, in
addition to the restoration, associated motion blur kernels,
segmentation, and motions are simultaneously obtained.
This algorithm has a few limitations too.
C. Paramanand and A. N. Rajagopalan [4] addressed the
problem of estimating the latent image of a static bilayer
scene (consisting of foreground and background of different
depths).They initially propose a method to estimate the
transformation spread function (TSF) corresponding to one
of the depth layers. The estimated TSF (which reveals the
camera motion during exposure) is used to segment the
scene into the foreground and background layers and
determine the relative depth value. The deblurred image of
the scene is finally estimated within a regularization
framework by accounting for blur variations due to camera
motion as well as depth.
ShamikTiwari, V. P. Shukla, A. K. Singh, S. R. Biradar [14]
described an estimate to the angle of blur kernel and length
of blur kernel. After estimation they used these estimated
parameters to deblur the image using various algorithms.
Hough transformation method, Radon transformation
method, Steerable filters method are used to estimate the
blur kernel and blur length.
Hanghang Tong, Mingjing Li, Changshui Zhang [6]
proposed a scheme that uses the harr wavelet transform
(HWT) in discriminating different types of edges as well as
recovering sharpness from the blurred version, and then