Issuu

International Research Journal of Engineering and Technology (IRJET)

e-ISSN: 2395-0056

Volume: 12 Issue: 05 | May 2025

p-ISSN: 2395-0072

www.irjet.net

REVIEW PAPER ON SMART FRUITS HARVESTING ROBOT Khushi Joshi1, Swastika Tripathi2, Anshika Patel3, Er.Sonam Singh4 *1234Department of Electronics And Communication Engineering SRMCEM, Lucknow, Uttar Pradesh, India

------------------------------------------------------------***-----------------------------------------------------------ABSTRACT The agricultural sector is increasingly turning to automation to address labor shortages and enhance productivity. One promising development in this area is the smart fruit-picking robot which helps to improve efficiency, reduce labor dependency, and maintain consistent quality. This project introduces a smart fruit harvesting robot specifically designed to efficiently harvest lightweight fruits such as oranges, apples, and guavas using advanced image processing techniques, Convolutional Neural Networks (CNNs) implemented in Python and the You Only Look Once (YOLO) object detection algorithm to identify and locate ripe fruits in real time. The system integrates a high-resolution camera with a Raspberry Pi to capture real-time images of the plants. Using Python, the images are processed to filter background noise and isolate potential fruit candidates based on color, shape, and texture. A pre-trained CNN model then classifies the fruits based on ripeness, ensuring only ripe fruits are selected for harvesting. Once a ripe fruit is identified and localized, the robot calculates its position using depth estimation and coordinates a robotic arm with a soft gripper to harvest the fruit gently and precisely. The system is equipped with basic obstacle avoidance and path-planning capabilities, allowing it to navigate through fields autonomously. Preliminary testing has shown high accuracy in fruit detection and effective harvesting with minimal damage to both the fruits and plants. The proposed smart fruit harvesting robot demonstrates significant potential in reducing labor costs, increasing harvesting efficiency, and contributing to the advancement of precision agriculture through AI-driven automation.

Keywords: Convolutional Neural Networks (CNN), Precision Agriculture, Robotic Arm with soft gripper, Autonomous Navigation.

INTRODUCTION

The agricultural industry is undergoing a technological transformation, with automation playing a key role in addressing challenges such as labor shortages, increasing demand, and the need for precision farming. One critical task in horticulture that demands precision and efficiency is fruit harvesting, particularly for lightweight and delicate fruits such as oranges, apples, guavas. Manual harvesting of these fruits is labor-intensive, time- consuming, and prone to inconsistency and damage. To overcome these challenges, smart robotic solutions are being developed to automate the fruit-picking process with higher speed, accuracy, and reliability. This project focuses on the development of a smart fruit harvesting robot that utilizes image processing techniques and artificial intelligence to detect and harvest lightweight fruits efficiently A traditional fruit harvesting relies heavily on manual labor, which can be time-consuming, costly, and prone to inconsistencies due to human error and fatigue. In contrast, smart fruit harvesting robots use advanced technologies such as computer vision, machine learning, robotic arms, and sensors to identify ripe fruits, determine the best picking strategy, and harvest without damaging the fruit or plant. A high-resolution camera captures images of the crops, which are processed using OpenCV in Python to enhance image quality and remove noise. YOLO quickly detects multiple fruits in a single frame, and a CNN model verifies their ripeness. As we refer to other models we found the arm mechanism becoming complex and not easy to function the robot. So we use scissor lifts technique for heightened tress to pick the fruits without causing harm and this will make the mechanism less complex , costeffective and effiecnt. Once identified, a robotic arm with a soft gripper mechanism is deployed to gently pick the fruit without causing damage. The system also includes path planning for obstacle avoidance capabilities. This smart harvesting robot represents a signficant step toward intelligent, automated, and scalable agricultural practices.

Impact Factor value: 8.315

ISO 9001:2008 Certified Journal

Page 606