High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Image Processing Evaluation: A Review

Author(s):

Priyanka , Doon Valley Institute of Engineering Technology, Karnal; Sarita Bajaj, Doon Valley Institute of Engineering Technology, Karnal

Keywords:

MATLAB, PSNR, MSE

Abstract

Image processing domain has been seeing lots of extensions and apprises since its conceptualization. It’s said nothing can match which is seen by natural eye retina but in order to capture the memories, scientific data images have been clicked and video have been created. But when it comes to data collection the quality of image being captured becomes of crucial importance and this has led to establishment of image processing field. It’s not only about image capturing but it involves lot of background like capturing the exact characteristics of image to convert it into a digital signal so that it can be transmitted and restored back retaining the same characteristic as that of original image. Its common phenomenon that whenever any data is being transmitted via any media, data is vulnerable to get noisy and loose its original characteristics. Our main attempt in this thesis is to bring the innovative concept into discussion for image processing and de-noising. For better understanding; correlation among two phenomenon has been explored like Enhanced Empirical Mode Decomposition and wavelet transform keeping different parameters like Mean square error & Peak signal to noise ratio under observation. In this thesis a concept of smart Empirical mode decomposition is used which is blend of Enhance Empirical Mode Decomposition and wavelet thresholding along with relative differentiation between different kinds of threholding analysis. In lieu of coming up with more possible combinations for best of the de-noising technique a new act threshold limitation has been put low pass components or the other way is to keep it intact before adding inverse DWT. Increasing one more step towards image reconstruction, algorithm has been tested with image blurring error and it has been noticed that the proposed scheme works more efficiently for image de-blurring then de-noising. In previous study it is always found that, there is a research gap to decide which IMFs must be used for reconstruction once all are extracted. Hurst exponent has been sued to choose potential IMF. The whole algorithm is implemented in MATLAB R2013a’s image processing toolbox along with graph visualization toolbox is used. Results are compared on the basis of peak signal to noise ratio (PSNR) and mean square error (MSE).

Other Details

Paper ID: IJSRDV4I40384
Published in: Volume : 4, Issue : 4
Publication Date: 01/07/2016
Page(s): 428-431

Article Preview

Download Article