Image classification for content-based indexing

  title={Image classification for content-based indexing},
  author={Aditya Vailaya and M{\'a}rio A. T. Figueiredo and Anil K. Jain and HongJiang Zhang},
  journal={IEEE transactions on image processing : a publication of the IEEE Signal Processing Society},
  volume={10 1},
Grouping images into (semantically) meaningful categories using low-level visual features is a challenging and important problem in content-based image retrieval. Using binary Bayesian classifiers, we attempt to capture high-level concepts from low-level image features under the constraint that the test image does belong to one of the classes. Specifically, we consider the hierarchical classification of vacation images; at the highest level, images are classified as indoor or outdoor; outdoor… 

A Comparative Study of Global and Local Feature Representations in Image Database Categorization

This paper aims at comparing each of these feature representation methods and their combinations by using a standard classifier over thirty categories and shows that the combined global and block-based feature representation performs the best.

A two level approach for scene recognition

  • Le LuK. ToyamaGregory Hager
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
This paper presents a stratified approach to both binary (outdoor-indoor) and multiple category of scene classification, which first learns mixture models for 20 basic classes of local image content based on color and texture information, and produces 20 probability density response maps indicating the likelihood that each image region was produced by each class.

Efficient image selection for concept learning

The presented framework uses unsupervised learning to organise images based on low-level similarity, in effort to assist a professional annotator in picking positive and negative samples for a given concept.

State of the art of content-based image classification

This paper reviewed some of the most commonly used image classification approaches and it was clear that approaches using intermediate semantic concepts were more appropriate.

Improved Content-Based Image Classification Using a Random Forest Classifier

This paper uses random forest classifier, capable for handle over-fitting situation, to overcome the generalization problem of classifier and extract the effective low-level colour and texture features from the images.

Efficient Image Classification on Vertically Decomposed Data

This paper proposes a nearest neighbor based algorithm in image classification that vertically decomposes image features into separate bit vectors, one for each bit position of the values in the features, and approximates a number of candidates of nearest neighbors by examining the absolute difference of total variation between the images in the repositories and the unclassified image.

Image Classification Using Fusion of Holistic Visual Descriptions

Two state-of- the-art soft computing machine learning techniques: Support vector machine (SVM) and Artificial Neural Networks have been used to classify scene images into meaningful categories to report satisfactory categorization performances on a large data set of eight categories.

An improved image classification based on K-means clustering and BoW model

A revised algorithm based on weighted visual dictionary of K-means cluster that is efficient in descript image information and can provide image classification performance and gets the visual dictionary.

An Efficient Image Retrieval Based on Fusion of Fast Features and Query Image Classification

An efficient and fast CBIR system is proposed, which is based on a fusion of computationally light weighted colour and texture features; chromaticity moment, colour percentile, and local binary pattern (LBP) and a supervised query image classification and retrieval model, which filters all irrelevant class images.



Content-based hierarchical classification of vacation images

A learning paradigm to incrementally train the classifiers as additional training samples become available is developed and preliminary results for feature size reduction using clustering techniques are shown.

Bayesian framework for semantic classification of outdoor vacation images

This paper casts the image classification problem in a Bayesian framework, and demonstrates how high-level concepts can be understood from specific low-level image features, under the constraint that the test images do belong to one of the delineated classes.

On image classification: city images vs. landscapes

Indoor-outdoor image classification

  • M. SzummerRosalind W. Picard
  • Computer Science
    Proceedings 1998 IEEE International Workshop on Content-Based Access of Image and Video Database
  • 1998
This work systematically studied the features of: histograms in the Ohta color space; multiresolution, simultaneous autoregressive model parameters; and coefficients of a shift-invariant DCT to show how high-level scene properties can be inferred from classification of low-level image features.

Color indexing with weak spatial constraints

This work proposes an approach that lies between uniformly tesselating the images with rectangular regions and relying on fully segmented images, and encoding a minimal amount of spatial information in the index to improve the discrimination power of color indexing techniques.

Relevance feedback: a power tool for interactive content-based image retrieval

A relevance feedback based interactive retrieval approach that effectively takes into account the subjectivity of human perception of visual content and the gap between high-level concepts and low-level features in CBIR.

NeTra: A toolbox for navigating large image databases

An implementation of NeTra, a prototype image retrieval system that uses color, texture, shape and spatial location information in segmented image regions to search and retrieve similar regions from the database, is presented.

Efficient and Effective Querying by Image Content

A new theorem is introduced that makes indexing possible by bounding the non-Euclidean, full cross-term quadratic distance expression with a simple Euclidean distance, and it is illustrated how orthogonal transforms, such as the Karhunen Loeve transform, can help reduce the dimensionality, without introducing false dismissals.

Comparing images using color coherence vectors

It is shown that CCV’s can give superior results to color histogram-based methods for comparing images that incorporates spatial information, and to whom correspondence should be addressed tograms for image retrieval.

Scheme for visual feature-based image indexing

This paper presents the initial work in developing an efficient indexing scheme using artificial neural network, which focuses on eliminating unlikely candidates rather than pin-pointing the targets directly in order to achieve content-based image indexing and retrieval.