Search

Abstract The quality and uniformity of green coffee beans significantly influence the overall flavor and value of the product. In the coffee industry, automated flaws and bean-type identification offer numerous advantages. This study examines the effectiveness of multiple YOLO (You Only Look Once) models for identifying and classifying green coffee beans. Various YOLO variants, including YOLOv3, YOLOv4, YOLOv5, YOLOv7, YOLOv8, and custom models, are compared with a focus on computational efficiency, accuracy, and speed. Utilizing a dataset of 4,032 training and 506 testing images encompassing diverse bean types, defects, and lighting conditions, we assessed the performance. The bounding boxes generated by our models accurately encompass coffee beans, with the background typically uniform and set to black. Our analysis reveals the superior performance of the custom-YOLOv8n model, which we meticulously customized for green coffee bean detection. This model achieved high precision, recall, f1-score, and mAP, demonstrating its potential for real-world implementation in coffee bean quality control systems. The customization process involved fine-tuning the model to focus on significant features relevant to green coffee bean detection and employing specific labeling strategies. Customization allows you to fine-tune the model to focus on important features relevant to green coffee bean detection. This sensitivity ensures that the model can effectively distinguish between different bean types and detect even subtle defects. This paper clarifies our primary objective of evaluating YOLO models’ performance in identifying and categorizing green coffee beans, with potential implications for enhancing efficiency and consistency in the coffee industry. A succinct key sentence underscores the benefits of our research for readers seeking efficient YOLO model selection and implementation in agricultural systems.

Abstract A reliable and accurate diagnosis and identification system is required to prevent and manage tea leaf diseases. Tea leaf diseases are detected manually, increasing time and affecting yield quality and productivity. This study aims to present an artificial intelligence-based solution to the problem of tea leaf disease detection by training the fastest single-stage object detection model, YOLOv7, on the diseased tea leaf dataset collected from four prominent tea gardens in Bangladesh. 4000 digital images of five types of leaf diseases are collected from these tea gardens, generating a manually annotated, data-augmented leaf disease image dataset. This study incorporates data augmentation approaches to solve the issue of insufficient sample sizes. The detection and identification results for the YOLOv7 approach are validated by prominent statistical metrics like detection accuracy, precision, recall, mAP value, and F1-score, which resulted in 97.3%, 96.7%, 96.4%, 98.2%, and 0.965, respectively. Experimental results demonstrate that YOLOv7 for tea leaf diseases in natural scene images is superior to existing target detection and identification networks, including CNN, Deep CNN, DNN, AX-Retina Net, improved DCNN, YOLOv5, and Multi-objective image segmentation. Hence, this study is expected to minimize the workload of entomologists and aid in the rapid identification and detection of tea leaf diseases, thus minimizing economic losses.