This document describes a content-based image retrieval system using features extracted from error diffusion block truncation coding (EDBTC). EDBTC produces color quantizers and a bitmap image, which are vector quantized to generate features for image matching - a color histogram feature and a bit pattern histogram feature. These features are used to measure similarity between a query image and images in a database. The proposed approach outperforms other content-based image retrieval methods on natural and textural image datasets. Hardware and software requirements for implementing the system are also provided, along with the project workflow.