Author Details
( * ) denotes Corresponding author
The primary goal of a content-based image retrieval (CBIR) system is to quickly find the appropriate images from a vast image collection for use in real-time applications. The creation of both prominent picture characteristics and small dimensional features is closely related to the performance and efficacy of the CBIR system, respectively. As a result, in this paper, we have performed image retrieval with small dimensional salient image components or features compared to the original image size, and the image retrieval accuracy has been improved due to the consideration of local information rather than global information of image data. Not all pixels are necessary for feature extraction, so the quantization approach is used to begin the operation. Since grayscale photographs lack color information, we have carefully retrieved texture and shape features while considering the image’s directionalities and geometry. LBP and GLCM are used to extract the texture, and EDH and adaptive tetrolet transformation are used to capture the geometry at various levels. Through numerous benchmark databases, this technique has been verified.
Keywords
Content-Based Image Retrieval (CBIR); EDH; LBP; GLCM; Tetrolet