Digital image processing

The digital image processing deals with developing a digital machine that plays operations on a digital image. It’s far a sub field of signals and systems but attention especially on pics. DIP focuses on developing a PC machine this is able to perform processing on an image. The enter of that gadget is a virtual picture and the device manner that photo the use of green algorithms, and gives a picture as an output. The most non unusual instance is adobe Photoshop. It far one of the widely used ability for processing digital photos. 

digital image processing

An image captured with the aid of a camera and sent to a virtual device to do away with all of the different details, and just recognition on the water drop by zooming it in this type of way that the satisfactory of the picture remains the same. Also the correct colorization and enhancement .of images play important role

Image enhancement:

Image enhancement in Digital Image Processing strategies are based totally on subjective photo exceptional criteria. No goal mathematical criteria used for optimizing processing outcomes. 

Brightness     J=1MNi=1MJ=1NF(i,j)

Contrast         C=1MNi=1Mj=1N[fi,j-j]2

image enhancement

Image histogram: 

Image histogram represents the statistical distribution of photo pixel brightness.

Image histogram

Image restoration:

Restoration in Digital Image Processing is a technique that tries to reconstruct or recover an photo that has been degraded  by using a prior expertise of the degradation phenomenon.

Restoration techniques:

  1. The inverse filters
  2. The wiener filter
  3. MAP formulation
  4. Median filter
  5. Adaptive filter
  6. Linear filter
  7. IBD method
  8. NAS-RIF
  9. Super resolution restoration algorithm based on gradient  adaptive interpolation
  10. Deconvolution using  a spare prior
  11. Block matching
  12. DRF
  13. Lucy Richard algorithm techniques

Median filter:

                This is the statistical approach as applied by the means of the name. On this method pixel fee changed by way of the median of the pixels in the neighborhood located.  The usage is to eliminate the salt and paper noise. It used broadly and can lessen the noise inside the photographs excellently. This filtering removes the noise however maintains the rims. This tends to conquer the photo to become blur and that is its advantage over the smoothing model

Adaptive filter:

          In adoptive filter behavior adjustments based on statistical characteristics of image within the filter area. It is that sort of linear filter out which has a transfer feature controlled with the aid of a variable parameter. For elimination of impulse noise in photos those filters use the color and gray area in contrast to other filters. It has pleasant noise suppression  consequences, hold edges in higher way and subsequently yield better first-rate .

Linear filter:

            In this, we replace every pixel with the aid of the linear aggregate of its neighbors. The operations which might carried out encompass polishing, smoothing and aspect enhancement.

Image degradation:

                  This technique given by via ayers and dainty (1988). Its miles a method of blind de-convolution. On this technique Fourier remodel calculated which causes much less computation. On this method photograph healing completed via earlier information of PSF. It effect in excessive decision and satisfactory. The drawback of this approach that convergence not assured.

The inverse filters:

Start from the generative model

g (u, v)= h (u ,v) *f (u, v)+n (u, v)↔ G(u, v) = H (u, v) * F (u, v) + N (u, v)

and for the moment ignore n (u, v) , then an estimate of f(u, v)  acquired form

 F(u, v) =G (u, v) / H (u, v)  


                    The filtering technique turned into recommend by way of D. Kunde.  The purpose is to reconstruct a reliable estimated image from a blurred photo. In this algorithm estimation of the target image is made. Errors characteristics are minimized to make the estimate which includes the domains name of photograph. The benefit is that we simplest want to locate aid domain of goal place need to be cautions such that the estimate obtained can be advantageous.

Super-resolution restoration algorithm based on gradient adaptive interpolation:

                    The primary idea is that the neighborhood gradient of pixel influences the interpolated pixel value. Inside the edge regions of the pixel. The influence is inversely proportional to the nearby gradient of a pixel. On this technique 3 subtasks are concerned: registration, fusion and blurring  

Deconvolution using a spare prior:

                     Deconvolution problem  is formulated in this set of rules as given by means  of the remark which determine the most a-posterior estimate of the unique photograph.

Block matching:

                    In block matching high correlation containing blocks are used because its accuracy is extensively laid low with the presence of noise. a block-similarity measure is applied which plays a rough preliminary de-noising in nearby second remodel area .

The wiener filter:

wiener filter


                  It is any other category of non-blind deconvolution approach. When smoothness like constraints are implemented on the picture recovered and restricted statistics of noise is known, then these strategies can be used effectively.

Lucy-Richardson algorithm techniques:

                   The photo recovery is split into blind non blind de convolution. In non-blind PSF is understood. The Lucy-Richardson is the maximum popular approach in the field of astronomy and medical imaging. The purpose of recognition is its potential to supply reconstructed pix of precise high-quality inside the presence.

MAP formulation:

         The MAP estimator has the form 

difference between image enhancement and image restoration

Difference between image enhancement and image restoration:


Read also about Visionary Interest: Why images are important.


Please enter your comment!
Please enter your name here

nine − 6 =