International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
Harvestify Dhruv Mahapure 1, Kunal Raut 2, Sumit Mishra 3, Maya Patil 4 1
Dhruv Mahapure, Student, Dept. of Information Technology, SJCEM , Palghar, Maharashtra, India 2 Kunal Raut, Student, Dept. of Information Technology, SJCEM , Palghar, Maharashtra, India 3 Sumit Mishra, Student, Dept. of Information Technology, SJCEM , Palghar, Maharashtra ,India 4 Maya Patil, Professor, Dept. of Information Engineering, SJCEM, Palghar, Maharashtra , India ---------------------------------------------------------------------***--------------------------------------------------------------------location for compost application, accounting for rainfall, Abstract
maintaining crop quality, and addressing the variability of factors even within different areas of the same field. These challenges, compounded during significant agricultural decision-making processes, often require consideration of multiple variables and metrics.
Agriculture serves as the primary source of livelihood for a significant portion of India’s population and is an integral part of the primary sector. The Harvesting System, a machine learning-driven solution, aims to enhance the efficiency of the harvesting process for farmers. Implementing an accurate vision system capable of real-time fruit classification and analysis is essential for ensuring the cost-effectiveness and productivity of harvesting robots.
To address these complexities, a program is proposed to assist farmers in enhancing their productivity by continuously monitoring agricultural fields. For instance, online weather data, including rainfall trends and soil parameters, can guide decisions on which crops are best suited for specific locations. This system introduces a desktop application that leverages data analysis techniques to predict the most profitable crop yields based on current climate and soil conditions.
In many regions of India, farmers face challenges in crop cultivation due to unfavorable climate conditions and poor soil quality. Additionally, the lack of readily accessible assistance to guide farmers in selecting the appropriate crops using modern technological advancements exacerbates the issue. Illiteracy further hampers farmers from leveraging advancements in agricultural science, often resulting in continued reliance on traditional farming methods. This can hinder achieving optimal yields. For example, crop failure may result from inadequate fertilization or unpredictable rainfall patterns.
Key Features of the System: 1. Crop Recommendation 2. Fertilizer Recommendation 3. Plant Disease Prediction Agriculture has been a cornerstone of Indian culture since ancient times. Historically, farmers cultivated crops on their own land, adapting practices to meet their needs. However, the advent of modern technologies and techniques has led to a gradual shift in agricultural practices. This shift, while beneficial in some respects, has also resulted in a growing reliance on hybrid and artificial products, which may compromise health. Moreover, a lack of awareness about proper crop cultivation timing and location has contributed to changes in seasonal climatic conditions, negatively impacting essential resources like soil, water, and air, ultimately leading to food insecurity.
The increasing availability of agricultural data offers immense potential for improving crop management but also poses challenges in its effective utilization. This study introduces the Harvestify Classifier, a machine learning-based framework designed to enhance agricultural decision-making. By accurately classifying crops and forecasting yields, the system leverages diverse datasets such as satellite images, soil health metrics, and climatic conditions. Advanced algorithms, including Random Forest and Support Vector Machines (SVM), are employed to ensure precise predictions and robust performance.
KEYWORDS
Supervised Learning: Creates mathematical models using labeled data, with inputs and corresponding desired outputs.
Machine Learning, Processing, Training, Testing, Predictive Model, Text Processing
1. INTRODUCTION
Unsupervised Learning: Builds models from datasets containing only inputs, with no labeled outputs.
According to the 2011 census, 118.6 million farmers in India rely on agriculture for their livelihood, highlighting its critical role in the country. Farmers face several challenges, such as understanding soil conditions, determining the right time and
Semi-Supervised Learning: Develops models from partially labeled data, where some input samples lack associated labels. This paper focuses on improving crop yields through various techniques, including fertilizer recommendations
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 1209