A Lightweight Apple Detection Method in Real Orchard Environments Based on Improved YOLO

This project focuses on the development of a lightweight apple detection system for orchard environments using advanced YOLO-based models YOLOv11, YOLOv12, and YOLOv26. The goal is to provide an efficient solution for detecting apples in natural orchard settings, overcoming challenges such as varying lighting conditions, cluttered backgrounds, and changing environmental factors. The system employs state-of-the-art deep learning algorithms for accurate and fast apple detection, ensuring minimal computational load, making it suitable for deployment in resource-constrained environments. The dataset used in the project contains images captured from orchards with different lighting conditions and apple positioning. Through the use of YOLO models, the system can successfully identify apples, contributing to more efficient orchard management practices. Developed using Flask for the backend and HTML, CSS, and JavaScript for the frontend, the system allows users to upload orchard images for apple detection. The proposed solution addresses key challenges in precision agriculture, such as automating the fruit detection process and improving efficiency in orchard operations. The work is expected to enhance automation, reduce labor costs, and facilitate the scaling of fruit detection systems in large agricultural environments. The future direction includes enhancing detection accuracy, optimizing the system's performance, and extending the detection capabilities to other fruits or objects in orchard settings.

Keywords: Apple detection, YOLO, object detection, orchard environment, deep learning, machine learning, Flask, precision agriculture, lightweight system, image classification.