Informatics Publishing Limited

S. Sasikala ¹, S. Thavamani ²

Affiliations
1 Sree Saraswathi Thyagaraja College, IN
2 Department of Computer Science, Sree Saraswathi Thyagaraja College, IN

Abstract

The ever increasing amount of multimedia data creates a need for new sophisticated methods to retrieve the information one is looking for. Especially for the visual content this is still an unsolved problem. Thus content-based image retrieval attracted many researchers of various fields in an effort to automate data analysis and indexing. Finally, keywords assignment is subjective to the person making it. Therefore, content-based image retrieval (CBIR) systems propose to index the media documents based on features extracted from their content rather than by textual annotations. For still images, these features can be color, shape, texture, objects layout, edge direction, etc. This paper focuses on using the spatial feature of Texture Primitive Feature and Statistical Shape Features for image retrieval. Based on the analysis of the statistical distribution of the texture primitive, the spatial distribution map of each feature is presented to describe the image texture information. The shape features suggested here are edge histograms and Fourier-transform-based features computed for an edge image in Cartesian and polar coordinate planes. Experimental result shows that the proposed image retrieval system results in better retrieval of images.

Keywords

Content-Based Image Retrieval (CBIR), Spatial Feature, Texture Primitive Feature, Statistical Shape Feature.

Full Text

     

M. Yuvaraja , S. Thavamani

Abstract

Sentiment Analysis (SA) performs on specific domain to achieve higher level of accuracy. Extracting the unstructured data sentimental analyses plays a major role. SA is mainly for automatically predict sentiment polarity of positive or negative aspects of data. Sentiment Analysis problem is machine learning problems which provide the outcome based of supervised and unsupervised methods using labeled and unlabeled data. By extracting the data from this cross domain many techniques were used. This paper provides survey on sentiment analysis of various techniques, methods, algorithm and tools of SA to adapt the data in source and target domain to extract the relevant knowledge.

Keywords

Sentiment Analyses (SA), Survey. Cross Domain Analyses, SA Techniques and SA Methods.

Full Text