Digital Image Processing

Nikita Jain ¹, Shiv Kumar ²

Affiliations
1 G.I.T., Jaipur (Raj), IN
2 ACEIT, Jaipur, IN

Abstract

Neural Networks have been developed rapidly during the recent few years and it is extensively applied for the enhancement of the digital image. An image enhancement is the process for improving the quality of digital image. A large number of conventional enhancement methods have been proposed and developed. The traditional methods are contrast enhancement, histogram equalization, edge sharpening, variety of filters etc. On the other hand, the back propagation algorithms with feed-forward networks, sigmoid functions, feed back networks are designed for enhancement in the digital image with high probability. The aim of this study is to reveal a comparison between conventional networks and Neural Networks. To accomplish this purpose, no of experiments has been conducted and examined. For this experimental analysis, statistical method has been used to classify and characterize the behavior of the images. Experiments on images are implemented to confirm the validity of the proposed analysis. One of the purpose of the study was to identify the main factor affecting the image and result are obtained were validated with existing techniques. This paper focuses on three popular features of image enhancement that are auto enhancement, face detection and edge detection.

Keywords

Conventional Networks, Edge Detection, Face Detection, Neural Networks.

Full Text

     

Shiv Kumar ¹, R. K. Singh ², Manish Srivastava ³, Dinesh Varshney ⁴, Vijay K. Chaudhari ¹

Affiliations
1 Department of Information Technology, Technocrat Institute of Technology, Bhopal (M.P.), IN
2 MATS University, Raipur (C.G.), IN
3 Department of Computer Science and Engineering, Guru Ghansidas Central University, Bilaspur (C.G.), IN
4 Multimedia Regional Centre, Madhya Pradesh Bhoj (Open) University, Khandwa Road Campus, Indore (M.P.), IN

Abstract

In this paper titled "A New Approach for Tree-Structured Wavelet Transform based Texture Retrieval Analysis (TRA) by Using MatLab" is define Wavelet transform-based texture analysis, as I found in the different research, uses sub-band energy values as features, but not the order of energy values. In fact, a textured image, after a wavelet decomposition, yields an energy distribution which can be rank ordered with respect to the sub-bands. It has been found that the combination of the sub-band energy value and its ranking order leads to a more efficient texture retrieval mechanism.

Keywords

MatLab7.0, Wavelet Toolbox, Image Processing Toolbox, Algorithm.

Full Text

     

Neetesh Gupta ¹, Niket Bhargava ², Md. Ilyas Khan ¹, Shiv Kumar ³, Bhupendra Verma ⁴

Affiliations
1 Department of Information Technology, Technocrat Institute of Technology, Bhopal (M.P.), IN
2 Department of Computer Science, Bansal Institute of Technology, Bhopal (M.P.), IN
3 Department of Information Technology, Technocrat Institute of Technology-Bhopal (M.P.), IN
4 Department of Computer Science, Technocrat Institute of Technology, Bhopal (M.P.), IN

Abstract

For the purpose of content-based image retrieval (CBIR) An up-to-date comparison of state-of-the-art low-level color and texture feature extraction approach is presented in this paper. The CBIR problem is identified by us because there is a need to search the huge databases having images efficiently and effectively. in this paper we suggest a color and texture feature extraction algorithms. Special attention is given for CBIR is the similarity measurement using correlation coefficient with distinct distance matrices properties. A New approach for image retrieval technique is proposed to improve retrieval performance, and reduce the extraction search times. Matching is performed between the test image and the object image and quality of matching is measured in terms of grading.

Keywords

Similarity, Measurement, CBIR, Low Level Feature Extraction, Correlation Coefficient.

Full Text