Rik Das

Performance comparison of feature vector extraction techniques in RGB color space using block truncation coding for content based image classification with discrete classifiers

Content based image classification is a vital component of machine learning and is attaining increasing importance in the field of image processing. This paper has carried out widespread comparison of block truncation coding based techniques for feature vector extraction of images which is a precursor of image classification. A new block truncation coding (BTC) based technique using even and odd image parts for feature vector extraction is also introduced to perform image classification. The performances of classifier algorithms are compared in Receiver Operating Characteristic (ROC) Space. Two different categories of classifiers viz. K Nearest Neighbor (KNN) Classifier and RIDOR Classifier are being used to observe the degree of classification for various techniques under six different feature vector extraction environments.

Performance Boost of Block Truncation Coding based Image Classification using Bit Plane Slicing

Image classification demands major attention with increasing volume of available image data. The paper has shown performance boosting of image classification after associating Bit Plane Slicing with Block Truncation Coding (BTC) for feature extraction. Here more significant bit planes were considered for extraction of feature vectors. RGB color space was considered to carry out the experimentation. A database of 900 images was used for evaluation purpose. KeywordsPlane Slicing, BTC, CBIC, RGB

A Novel Feature Extraction Technique Using Binarization of Bit Planes for Content Based Image Classification

A number of techniques have been proposed earlier for feature extraction using image binarization. Efficiency of the techniques was dependent on proper threshold selection for the binarization method. In this paper, a new feature extraction technique using image binarization has been proposed. The technique has binarized the significant bit planes of an image by selecting local thresholds. The proposed algorithm has been tested on a public dataset and has been compared with existing widely used techniques using binarization for extraction of features. It has been inferred that the proposed method has outclassed all the existing techniques and has shown consistent classification performance.

Image Classification using Block Truncation Coding with Assorted Color Spaces

The paper portrays comprehensive performance comparison of image classification techniques using block truncation coding (BTC) with assorted color spaces. Overall six color spaces have been explored which includes RGB color space for applying BTC to figure out the feature vector in Content Based Image Classification (CBIC) techniques. A generic database with 900 images having 100 images per category spread across 9 different categories have been considered to conduct the experimentation with the proposed Image Classification technique. On the whole nine hundred queries have been fired. The average success rate of class determination for each of the color spaces has been computed and considered for performance analysis. The results explicitly reveal performance improvement (higher average success rate values) with proposed colorBTC methods with luminance chromaticity color spaces compared to RGB color space. Best result is shown by YUV color space based BTC in content based image classification.

Feature Extraction with Ordered Mean Values for Content Based Image Classification

Categorization of images into meaningful classes by efficient extraction of feature vectors from image datasets has been dependent on feature selection techniques. Traditionally, feature vector extraction has been carried out using different methods of image binarization done with selection of global, local, or mean threshold. This paper has proposed a novel technique for feature extraction based on ordered mean values. The proposed technique was combined with feature extraction using discrete sine transform (DST) for better classification results using multitechnique fusion. The novel methodology was compared to the traditional techniques used for feature extraction for content based image classification. Three benchmark datasets, namely, Wang dataset, Oliva and Torralba (OT-Scene) dataset, and Caltech dataset, were used for evaluation purpose. Performance measure after evaluation has evidently revealed the superiority of the proposed fusion technique with ordered mean values and discrete sine transform over the popular approaches of single view feature extraction methodologies for classification.

Retrieval architecture with classified query for content based image recognition

The consumer behavior has been observed to be largely influenced by image data with increasing familiarity of smart phones and World Wide Web. Traditional technique of browsing through product varieties in the Internet with text keywords has been gradually replaced by the easy accessible image data. The importance of image data has portrayed a steady growth in application orientation for business domain with the advent of different image capturing devices and social media. The paper has described a methodology of feature extraction by image binarization technique for enhancing identification and retrieval of information using content based image recognition. The proposed algorithm was tested on two public datasets, namely, Wang dataset and Oliva and Torralba (OT-Scene) dataset with 3688 images on the whole. It has outclassed the state-of-the-art techniques in performance measure and has shown statistical significance.

Novel Technique in Block Truncation Coding Based Feature Extraction for Content Based Image Identification

Feature vector extraction has been the key component to define the success rate for content based image recognition. Block truncation coding is a simple technique which has facilitated various methods for effective feature vector extraction for content based image recognition. A new technique named Sorted Block Truncation Coding (SBTC) has been introduced in this work. Three different public datasets namely Wang Dataset, Oliva and Torralba (OT-Scene) Dataset and Caltech Dataset consisting of 6,221 images on the whole was considered for evaluation purpose. The technique has stimulated superior performance in image recognition when compared to classification and retrieval results with other existing techniques of feature extraction. The technique was also evaluated in lossy compression domain for the test images. Various parameters like precision, recall, misclassification rate and F1 score has been considered to evaluate the performances. Statistical evaluations have been carried out for all the comparisons by introducing paired t test to establish the significance of the findings. Classification and retrieval with proposed technique has shown a minimum of 14.4 % rise in precision results compared to the existing state-of-the art techniques.

Partition selection with sparse autoencoders for content based image classification

Managing colossal image datasets with large dimensional hand-crafted features is no more feasible in most of the cases. Content based image classification (CBIC) of these large image datasets calls for the need of dimensionality reduction of features extracted for the purpose. This paper identifies the escalating challenges in the discussed domain and introduces a technique of feature dimension reduction by means of identifying region of interest in a given image with the use of reconstruction errors computed by sparse autoencoders. The automated process identifies the significant regions in an image for feature extraction. It not only improves the dimension of useful features but also contributes to increased classification results compared to earlier approaches. The reduction in number of one kind of features easily makes space for the inclusion of other features whose fusion facilitates improved classification performance compared to individual feature extraction techniques. Two different datasets, i.e. Wang dataset and Corel 5K dataset have been used for the experiments. State-of-the-art classifiers, i.e. Support Vector Machine and Extreme Learning Machine are used for CBIC. The proposed techniques are evaluated and compared in the context of both the classifiers and analysis of results suggests the appropriateness of the proposed methods for real time applications.

Decision fusion-based approach for content-based image classification

Purpose Current practices in data classification and retrieval have experienced a surge in the use of multimedia content. Identification of desired information from the huge image databases has been facing increased complexities for designing an efficient feature extraction process. Conventional approaches of image classification with text-based image annotation have faced assorted limitations due to erroneous interpretation of vocabulary and huge time consumption involved due to manual annotation. Content-based image recognition has emerged as an alternative to combat the aforesaid limitations. However, exploring rich feature content in an image with a single technique has lesser probability of extract meaningful signatures compared to multi-technique feature extraction. Therefore, the purpose of this paper is to explore the possibilities of enhanced content-based image recognition by fusion of classification decision obtained using diverse feature extraction techniques. Design/methodology/approach Three novel techniques of feature extraction have been introduced in this paper and have been tested with four different classifiers individually. The four classifiers used for performance testing were K nearest neighbor (KNN) classifier, RIDOR classifier, artificial neural network classifier and support vector machine classifier. Thereafter, classification decisions obtained using KNN classifier for different feature extraction techniques have been integrated by Z-score normalization and feature scaling to create fusion-based framework of image recognition. It has been followed by the introduction of a fusion-based retrieval model to validate the retrieval performance with classified query. Earlier works on content-based image identification have adopted fusion-based approach. However, to the best of the authors’ knowledge, fusion-based query classification has been addressed for the first time as a precursor of retrieval in this work. Findings The proposed fusion techniques have successfully outclassed the state-of-the-art techniques in classification and retrieval performances. Four public data sets, namely, Wang data set, Oliva and Torralba (OT-scene) data set, Corel data set and Caltech data set comprising of 22,615 images on the whole are used for the evaluation purpose. Originality/value To the best of the authors’ knowledge, fusion-based query classification has been addressed for the first time as a precursor of retrieval in this work. The novel idea of exploring rich image features by fusion of multiple feature extraction techniques has also encouraged further research on dimensionality reduction of feature vectors for enhanced classification results.

Decision Fusion for Classification of Content Based Image Data

Information recognition by means of content based image identification has emerged as a prospective alternative to recognize semantically analogous images from huge image repositories. Critical success factor for content based recognition process has been reliant on efficient feature vector extraction from images. The paper has introduced two novel techniques of feature extraction based on image binarization and Vector Quantization respectively. The techniques were implemented to extract feature vectors from three public datasets namely Wang dataset, Oliva and Torralba OT-Scene dataset and Corel dataset comprising of 14,488 images on the whole. The classification decisions with multi domain features were standardized with Z score normalization for fusion based identification approach. Average increase of 30.71% and 28.78% in precision were observed for classification and retrieval respectively when the proposed methodology was compared to state-of-the art techniques.