Content-Based Image Retrieval for Remote Sensing Images using Hybrid Firefly and Grey Wolf Optimization Algorithm

In recent years, Content-Based Image Retrieval (CBIR) has rised as an advanced approach for identifying and retrieving images from a database based on their visual content. CBIR analyzes visual material primarily using low-level properties such as color, texture, and shape. These qualities, however, may fall short in terms of capturing high-level semantics and contextual information, limiting the system's ability to fully understand the whole significance of images. To address this limitation, the hybrid Firefly and Grey Wolf Optimization (FGWO) algorithm has been proposed, which combines the Grey Wolf Optimization (GWO) with the Firefly Algorithm (FA). The FA is a nature-inspired swarm intelligence-based optimization method inspired by firefly flashing light behavior. This integration improves the image retrieval capabilities of the system. Furthermore, an Adjusted Intensity Based Variant of Adaptive Histogram Equalization(AIBVAHE) filter is utilized to efficiently reduce noise in satellite images. The AIBVAHE filter improves visual contrast by redistributing intensity levels, resulting in a more uniform histogram. Color Moments and Local Binary Pattern (LBP) are utilized as feature extractors in the hybrid feature extraction approach, efficiently extracting the most relevant features suited for Content-Based Satellite Image Retrieval (CBSIR) systems. When compared to existing methods, the experimental findings show that this methodology is successful, with an impressive accuracy of 98.91%, precision of 98.41%, and recall of 98.50%.