Raw Coffee Bean Classification for Roasting Suitability Assessment Using Transfer Learning

The classification of raw robusta coffee beans is a pivotal process with profound implications for the coffee industry. Recognizing the critical significance of this classification, this research endeavors to establish an effective and automated method for discerning coffee bean quality. The primary objective of this study is to employ advanced machine learning techniques to classify raw robusta coffee beans accurately. Specifically, we utilize deep learning models, VGG-16 and Inception V3, for this purpose. These models are trained to classify coffee beans based on their quality attributes, offering a systematic approach to segregating beans into high and low-quality categories. The research methodology encompasses comprehensive data preprocessing, model construction, and rigorous experimentation. Notably, the Inception V3 model, initialized with pre-trained weights from ImageNet, emerged as the top performer, achieving an outstanding accuracy rate of 99%. The implications of this research extend to various stakeholders within the coffee industry. By enhancing the quality assessment of raw robusta coffee beans, this research contributes to improved quality control processes. It ensures that only the finest beans proceed through the production chain, bolstering economic viability and market competitiveness for coffee-producing regions.