Facial Image Noise Removal Via a Trained Dictionary
Journal Title: INTERNATIONAL JOURNAL OF COMPUTER TRENDS & TECHNOLOGY - Year 2014, Vol 9, Issue 1
Abstract
In this project we address that, sparsity has shown to be useful in source separation. In most cases, the sources are not sparse currently and needs to sparsify them using a known dictionary. The problem here is that, if the sparse domain is not available then it will be difficult to recover the source using the current algorithms. In-order to address this problem we fuse the dictionary into the source separation. We define a cost function based on the idea and propose by extending the de-nosing method and minimize it. The term sparse refers to signals or images with small number of non - zeros with respect to some representation bases. In sparse component analysis (SCA), the assumption is that the sources can be sparsely represented using a known common basis or dictionary. The existing system defines that the techniques like MCA which is used provide a noisy mixture and present the source. In the proposed technique FastICA algorithm which employs a modified Gaussian for blind source separation. The proposed non-linear function which is used to separate image mixtures and result in faster execution and in good quality image separation
Authors and Affiliations
Sudharson. D , Kavinraj. A. S , Sridhar. S , C. Dinesh Kumar
Keywords
Related Articles
- EP ID EP152382
- DOI -
- Views 96
- Downloads 0
Analysis of Email Fraud Detection Using WEKA Tool
Data mining is also being useful to give solutions for invasion finding and auditing. While data mining has several applications in protection, there are also serious privacy fears. Because of email mining, even inexperi...
Vector Space Models to Classify Arabic Text
Text classification is one of the most important tasks in data mining. This paper investigates different variations of vector space models (VSMs) using KNN algorithm. The bases of our comparison are the most popular text...
Vertical Clustering of 3D Elliptical Helical Data
This research proposes an effective vertical clustering strategy of 3D data in an elliptical helical shape based on 2D geometry. The clustering object is an elliptical cross-sectioned metal pipe which is been bended in t...
Ensuring Accountability for Data Sharing in Cloud
Cloud computing is emerging rapidly as part of latest developments in communication trends and technology. It brought revolution in today’s world among key stakeholders like service providers and consumers. As a result,...
A Study on Educational Data Mining
Educational Data Mining is an emerging discipline, concerned with developing methods for exploring the unique types of data that come from educational settings, and using those methods to better understand students, and...