Query by Image Content using Color-Texture Features Extracted from Haar Wavelet Pyramid

Content based image retrieval (CBIR), is an application of real-world computer vision domain where from a query image, similar images are searched from the database. The research presented in this paper aims to find out best features and classification model for optimum results for CBIR system.Five different set of feature combinations in two different color domains (i.e., RGB & HSV) are compared and evaluated using Neural Network Classifier, where best results obtained are 88.2% in terms of classifier accuracy. Color moments feature used comprises of: Mean, Standard Deviation,Kurtosis and Skewness. Histogram features is calculated via 10 probability bins. Wang-1k dataset is used to evaluate the CBIR system performance for image retrieval.Research concludes that integrated multi-level 3D color-texture feature yields most accurate results and also performs better in comparison to individually computed color and texture features.

Download Full-text

Haar Wavelet Pyramid-Based Melanoma Skin Cancer Identification With Ensemble of Machine Learning Algorithms

International Journal of Healthcare Information Systems and Informatics ◽

10.4018/ijhisi.20211001.oa24 ◽

2021 ◽

Vol 16 (4) ◽

pp. 1-15

Author(s):

Sudeep D. Thepade ◽

Gaurav Ramnani

Keyword(s):

Machine Learning ◽

Skin Cancer ◽

Health Informatics ◽

Performance Metrics ◽

Learning Algorithms ◽

Haar Wavelet ◽

Machine Learning Algorithms ◽

Computer Assisted ◽

Marginal Improvement ◽

Wavelet Pyramid

Melanoma is a mortal type of skin cancer. Early detection of melanoma significantly improves the patient’s chances of survival. Detection of melanoma at an early juncture demands expert doctors. The scarcity of such expert doctors is a major issue with healthcare systems globally. Computer-assisted diagnostics may prove helpful in this case. This paper proposes a health informatics system for melanoma identification using machine learning with dermoscopy skin images. In the proposed method, the features of dermoscopy skin images are extracted using the Haar wavelet pyramid various levels. These features are employed to train machine learning algorithms and ensembles for melanoma identification. The consideration of higher levels of Haar Wavelet Pyramid helps speed up the identification process. It is observed that the performance gradually improves from the Haar wavelet pyramid level 4x4 to 16x16, and shows marginal improvement further. The ensembles of machine learning algorithms have shown a boost in performance metrics compared to the use of individual machine learning algorithms.

Download Full-text

Broccoli Seedling Segmentation Based on Support Vector Machine Combined With Color Texture Features

IEEE Access ◽

10.1109/access.2019.2954587 ◽

2019 ◽

Vol 7 ◽

pp. 168565-168574

Author(s):

Kunlin Zou ◽

Luzhen Ge ◽

Chunlong Zhang ◽

Ting Yuan ◽

Wei Li

Keyword(s):

Support Vector Machine ◽

Texture Features ◽

Support Vector ◽

Color Texture

Download Full-text

Indexing of Image Features Using Quadtree

Emerging Technologies in Intelligent Applications for Image and Video Processing - Advances in Computational Intelligence and Robotics ◽

10.4018/978-1-4666-9685-3.ch008 ◽

2016 ◽

pp. 185-203

Author(s):

N. Puviarasan ◽

R. Bhavani

Keyword(s):

Extraction Method ◽

Dimensional Space ◽

Texture Features ◽

Image Features ◽

High Dimensional ◽

Shape Features ◽

Indexing Methods ◽

Feature Extraction Method ◽

Indexing Method ◽

Color Texture

In Content based image retrieval (CBIR) applications, the idea of indexing is mapping the extracted descriptors from images into a high-dimensional space. In this paper, visual features like color, texture and shape are considered. The color features are extracted using color coherence vector (CCV), texture features are obtained from Segmentation based Fractal Texture Analysis (SFTA). The shape features of an image are extracted using the Fourier Descriptors (FD) which is the contour based feature extraction method. All features of an image are then combined. After combining the color, texture and shape features using appropriate weights, the quadtree is used for indexing the images. It is experimentally found that the proposed indexing method using quadtree gives better performance than the other existing indexing methods.

Download Full-text