Authors: Arpitha I M, A ThyagarajaMurthy
Certificate: View Certificate
Agriculture is the backbone of our country and contributes almost 18 percent of the GDP, according to the Department of Agriculture, Cooperation, and Farmer Welfare’s 2018–2019 annual report. Agriculture is the largest sector of the Indian economy, employing about half of the labour force. However, farmers might not always benefit from the market for their products. Producing the fruits and vegetables, India is second largest in world according to Department of Agriculture, cooperatives and Farmer Welfare’s annual report for 2020–2021 yet, farmers there are experiencing financial hardship due to crop failures. Marketing veggies from growers to consumers is fraught with issues. Transportation and the seasonality of fruit are the two rather significant obstacles that farmers who grow vegetables must overcome. The perishability of the goods, the seasonality of manufacturing, the bulkiness of the products, and the price difference between buying and selling were the less severe constraints. The objective of this project is to create a system that helps farmers choose the best crop based on soil characteristics, environmental factors like temperature, soil humidity, humidity and market demand, as well as technological advancements like crop suggestions and decision-making models to help farmers understand the demand for perishable crops.
If cultivation is put down and distant from the corn field, ploughing may be simpler. Starting with our progenitors, the FERTILE CRESCENT, it all began in the globe around 10,000B.C. Hunters, who were nomadic people that moved from place to place in quest of sustenance, eventually started gathering wild gourds that they discovered growing on the soil. The extra grain was then dispersed to help grow additional food. But today, there is no food crisis, no agricultural development, no use of pesticides, and no need for specialised farming machinery. Things are much better and very different than they were in the past. They nevertheless lead a healthy lifestyle, are free from marketing disputes, and do not had any farmer suicides. This is due to the fact that the agricultural land was fruitful and overfilling the land was not a problem at the time. The current state of agriculture deviates from what farmers anticipated. Although agriculture in our country has improved, the yield of our main horticultural and agricultural crops is still quite poor when compared to other nations. Our agricultural sector still lacks in technology. In our country, food grain, fruit, and vegetable yields per hectare are much below average globally. Our rice yield is roughly half that of Vietnam and Indonesia, and it is only a third of China’s. Even the most productive regions of India under perform the world. Similar to this, by paying attention to seeds, soil health, pest management, crop-saving irrigation, and post-harvest technologies, the production of pulses and oil seeds may be boosted by 2.3 to 2.5 times. By 2025, India’s population is predicted to exceed 1.5 billion, making food security the most pressing societal issue and necessitating a significant increase in food production to accommodate the expanding population. A quarter of all undernourished individuals worldwide, or 217 million people, live in India, according to the FAO. To increase agricultural productivity, it is necessary to adopt new technologies including biotechnology, nanotechnology, high-tech protected cultivation, and contemporary irrigation techniques.
A. Fruits and Vegetables Sector Scenerio
As of the end of 2019 (NHB, 2019), India produced a total of 88.97 million metric tonnes of fruits and 162.89 million metric tonnes of vegetables, placing it second only to China in terms of fresh agro-food output. The proportion of fruits increased from the 2019–2020 period to 88.97 million metric tonnes, while the proportion of vegetables increased from 88.62 million metric tonnes to 162.89 million metric tonnes. Despite being the second-largest producer of fruits and vegetables in the world, supply chain losses and waste make it difficult for customers to get the ideal items at the ideal time and price. India is one of the top wasters in the world due to a significant portion of its production going to waste. It is also discussed how the very inefficient fruit and vegetable supply chain in India results in significant losses and waste, as well as lower profits for stakeholders.
In addition to the farmers losing money, it increases other costs across the supply chain, forcing the final customer to pay exorbitant prices out of his own pocket. Numerous investigations of postharvest losses in India have unequivocally demonstrated that the quantity of food wasted there annually is equal to the quantity consumed there. According to a recent assessment conducted in India, fruit and vegetable producers lose out on an estimated Rs. 2.13 lakh crore annually as a result of losses in the supply chain.
C. Measures for Improving Supply Chain of Fruits and Vegetables and its Effectiveness
Objective of the Proposed system are:
II. LITERATURE SURVEY
Middle-class people are more profitable than farmers as the link between producer and customer; this is especially true in agricultural marketing. As a result, a lot of money is being lost by farmers. To tackle this, developers have built a mobile and internet application that enables farmers to sell their products to clients..
Due to social media, eroding relationship ideals, demanding schedules, and possibly a wide range of possibilities, some wish they were alone in today’s culture. Sometimes, though, all they want are close friends who they can confide their feelings in. To address these concerns, they have developed a smartphone app that enables users to find their closest friends and invite them to join them. This three-layer structure is composed of a website as the third layer, an event controller and manager as the second layer, and user interaction as the first layer..
A client-client application using ASP.NET was created to provide good fruit quality to end users and to obstruct communication between producer and customer due to the overall increased productivity of fruits and vegetables. This makes it possible for customers to buy fruit straight from the grower..
The final user can define the kind of crop they need and how much they need (in kg). Farmers will evaluate the crop kind, the number of hectares they have planted, and the fair price. The programme will give a list of farmers who meet those criteria by evaluating the needs of end customers. Then, if he decides to purchase the crop, he can send the farmer an application so that they can communicate..
One of the most important facts is that transportation is crucial to advertising, and for that to happen, the roads need to be in decent condition. Additionally, this paper outlines a method for finding a quick and safe route in the presence of damage or impediments. By measuring the length and width of the roadways, a certain remote sensor image can be precisely removed from the road network (both damaged and good).
Road instability is found using multi-level search by repeatedly increasing the jump point by pixels (caused by damage or blockage). The test results demonstrate the accuracy and efficiency of the suggested procedure..
This model is recommended to implement a chain of food supply using blockchain technology to prevent third parties from entering the market, as well as market intelligence using building blocks of big data analytics and an all-in-one mobile application to assist farmers for an easy accessibility so that farmers register to use facilities by providing personal details and adhar services. For their farm, they might also oversee various food chain operations and keep an eye on recent market trends, weather predictions, projections, and neighbouring market locations..
In reaction to meteorological and geographic changes, this programme, which was developed using a machine learning algorithm, compels farmers to choose what crops to produce. They have also developed a secondary model that forecasts rainfall for the coming 12 months in order to help farmers..
For farmers, a crop suggestion and prediction model based on variables including humidity, rainfall, and temperature was developed. It makes use of collaborative filtering, multi-condition collaborative algorithms, Naive Bayes, KNN, SVMN, random foresight, and SVMN. It classifies the crop according to the user-provided high, low, and moderate parameter ranges; it then lists the top 5 crops in order; and it displays the crop..
A virtual simulation is run by analysing how the government, the farmer, and the consumer make decisions. In order to increase the safety of agricultural goods and attain food safety, this is done by analysing behaviour to an ideal state. Proper monitoring may also persuade the farmer to switch from using non-green pesticides to employing green pesticides..
A web application has been developed to help farmers get a fair price. Through it, an administrator can post an image of agricultural products with information from the seller, users can log in and place bids on the product, the seller sets the bidding period, and at the end of the bidding, a fiercely competitive consumer purchases the product, cutting out the middleman.
The paper offers details on how blockchain technology is applied and how it is used in the farming sector. The technical elements were well documented, including data structures, cryptographic methods, etc. After that, an assessment of all blockchain technologies was conducted in the second sector to clarify how they were used. Additionally, illustrations of typical techniques were provided to illustrate how they could be used to develop agricultural applications. The third section identifies the challenges that various facets of the agriculture system face to aid in understanding..
By offering an excellent agriculturist certificate based on the previous crops that they have grown, the model is introduced with the help of the leaders of the farmers union and businesspeople, allowing the farmers to improve the quality of their production and prevent a loss of agricultural products from farmers who grow the crop without knowing the demand for the crop, without a proper plan, without checking a current market trend, or climatic change..
To address this, they used Python to construct the application, allowing them to communicate with consumers directly and offer advice, notably regarding what crop should be planted when based on various criteria, as well as supply farmers with notifications of popular crops. In the current world, brokers are increasing their profits by preying on farmers’ ignorance as they toil round-the-clock to cultivate a crop..
In order to analyse the effects of mobile technology, this research largely focuses on how it is used. By questioning farmers about the daily applications they use and how much time they spend on them, researchers have also looked at how farmers communicate. .
Price forecasting has become a significant agricultural problem in the current era that can only be solved with the facts at hand. The objective of this article is to forecast crop price for the upcoming rotation. The goal of this research is to identify suitable data models that provide high accuracy and universality for price prediction. A variety of data mining techniques were investigated on various data sets to address this issue.This article describes a system that estimates agricultural prices using data analytics techniques. The suggested approach would forecast agricultural prices based on a number of factors, including crop area harvested and area seeded, among others, using machine learning algorithms.
A farmer can also get an idea of the price of the produce he will gather in the future. The goal of the study is to create a system that can foresee the crop’s target price and help farmers over the long term by combining data from various sources, data analytics, and prediction analysis. The overall study’s findings suggest that the optimum approach for this project is XGBoost..
The agricultural industry has a considerable impact on the Indian economy. The bulk of Indians, either openly or surreptitiously, depend on agriculture for their livelihood.
Agriculture’s importance to the country is therefore undeniable. The vast majority of Indian farmers feel that choosing which crop to sow during a specific season should be left to their senses. They find comfort in just following the patterns and customs of traditional farming, refusing to acknowledge that crop productivity is dependent on the current weather and soil conditions. However, it is unrealistic to expect a single farmer to take into account all the various factors that affect crop development when choosing which crop to grow.
A farmer could make a hasty or negligent judgement that has unfavourable effects on both him and the local agricultural sector. Machine learning and big data analytics can work together to solve this problem. In the work, present the AgroConsultant intelligent system, which intends to assist Indian farmers in selecting the optimum crop based on the sowing season, the location of their farms, the characteristics of the soil, and environmental factors like temperature and rainfall..
study compared and contrasted the conventional statistical maximum likelihood method with random forests and support vector machines using 126 features from Sentinel-2A images. In 2017, Sentinel-2A photos were successfully used to extract spectral reflectance of 12 bands, 96 textural features, 7 vegetation indicators, and 11 phenological parameters. According to the classification results, when 13 features are correctly combined, both traditional classification and machine learning obtain total classification accuracy of 88.96 percent and 98.8 percent, respectively. Short-wave infrared data can be used to identify rice, corn and soybeans significantly. Water vapour band can be used to distinguish between rice and corn. While the conventional classification result shows uneven accuracies for different crops, machine learning approaches show resilience with identification accuracy of greater than 95 percent for each crop variety. .
Agriculture is essential to human existence. It is one of the primary sources of employment in India. More over half the population is supported by the agriculture sector. It serves as the cornerstone of our economy. Crop productivity is affected by numerous factors. One of the main factors affecting crop productivity is soil. By improving the techniques used to forecast agricultural yield under varied climatic conditions, farmers and other stakeholders may be able to make better agronomic decisions. Crop yield prediction is process of estimating a crop’s production using historical information, which may include elements like temperature, humidity, pH, rainfall and crop’s name. It provides us with a general concept of the best crop that can be grown under the field weather conditions at the moment. Machine learning algorithms have been employed in the proposed study to compare soil properties and forecast crop production and fertility. To estimate agricultural yield, the Decision Trees classifier, Nave Bayes algorithm, and K Nearest Neighbor algorithm were utilised..
Teaching computers how to use data or past knowledge to solve problems in the real world is the core goal of machine learning. Through supervised learning, unsupervised learning, and reinforcement learning, it can be used as association analysis. In this investigation, we will concentrate on applying machine learning to pertinent issues in agriculture. We’ll concentrate on innovative uses of machine learning in agriculture. We also investigate using machine learning to grow wheat. Finally, we pinpoint the production gaps in wheat harvests and present a supervised machine learning-based method..
The Hellenic Agricultural Organization constantly supported Ecodevelopment S.A.’s RD activities, which were focused on offering precision farming services to rice producers and produced this study (Demeter). Within this framework, a novel machine learning-based topdressing nitrogen prediction method was developed. Nitrogen is a key component of rice cultivation and can be wisely controlled to increase yield, save costs, and protect the environment. Thessaloniki Plain in Greece’s 110 ha experimental rice field was continuously monitored for four years, resulting in a multi-source, multi-temporal, and multi-scale dataset that included optical and radar pictures, soil data, and yield maps. In order to identify waterlogged fields and, consequently, to pinpoint the specific growth stage of the crop, Utilizing computer vision, Sentinel-1 (SAR) imagery was simulated. Before applying topdressing treatments, leaf nitrogen concentration (LNC) was precisely mapped using Sentinel-2 image data. To define management zones, RapidEye imagery underwent image segmentation. To estimate yield for different nitrogen levels, a number of machine learning techniques were used, with the XGBoost model showing the highest accuracy. Utilising yield curves, the nitrogen dose that would maximise production was selected, and the farmer was counselled to use that amount. Inundation mapping ended up being a huge help to the prediction strategy. Ecodevelopment S.A. is currently expanding the new method’s application over a wide range of topic areas in an effort to boost its applicability and operationality..
The agricultural industry has a considerable impact on the Indian economy. The bulk of Indians, either openly or surreptitiously, depend on agriculture for their livelihood. Agriculture’s importance to the country is therefore undeniable. The vast majority of Indian farmers feel that choosing which crop to sow during a specific season should be left to their senses. They find comfort in just following the patterns and customs of traditional farming, refusing to acknowledge that crop productivity is dependent on the current weather and soil conditions. However, it is unrealistic to expect a single farmer to take into account all the various factors that affect crop development when choosing which crop to grow. The farmer can suffer negative consequences as a result of one rash or unwise decision, which would affect both him and the local agricultural sector. To overcome this issue, big data analytics and machine learning complement one other nicely. In this article, we introduce AgroConsultant, an intelligent system that will assist Indian farmers in selecting the optimum crop for their land based on the sowing season, their farm’s location, the soil’s characteristics, and weather factors like temperature and rainfall..
Climate change, diminishing productivity, and a lack of knowledge of agricultural innovation are all current issues in agriculture. One of the main factors contributing to low agricultural production and the underdeveloped state of agriculture is the division, slicing, and segmentation of land holdings. Moving cattle, manure, seeds, and pesticides from one piece of land to another costs farmers a lot of time and energy.Agriculture is significantly impacted by a change in the climate. In this instance, greater temperatures discourage the growth of pests and weeds while lowering crop yields to their ideal levels. Changes in rainfall and precipitation patterns raise the possibility of both short-term crop failures and long-term output declines. We proposed a method for crop suggestions that bases its forecasts on elements such soil type, weather conditions, and the location of agricultural land. We also develop a platform where farmers may list vacant land for cultivation and get agricultural land to lease in exchange. Through the news feed interface, users of this programme can learn about the most recent advancements in agriculture. search the appropriate location and select Agri assistance..
A. Existing Technologies
B. Drawbacks of Existing Models
As we’ve seen, farmers’ aid has a variety of uses, including marketing, crop suggestions, crop yields, crop price forecasts, and more. However, farmers utilise these types to predict their yield, earnings, markets for their crops, and what they can grow in their fields. However, there is no set methodology for guiding farmers prior to planting in order to dispel the misconception of overproduction of the same crops. And there isn’t a perfect decision-making tool that can help farmers decide how much of each crop to plant where and how much of a given crop to produce.
III. PROPOSED SYSTEM
This Section shows the methodology used in the work: The application is based on the three construction phases mentioned below:
B. Design Modules
It is possible to download this decision-making model for Android. Farmers enter their username, phone number, and password while registering for the application. They will then enter their login information, upload crop information, and have this app download the farmer’s current location. This will allow them to check other farmers’ crop information and determine whether a market exists for the product they have just launched. they develop, have the ability to choose whether or not to plant crops, and can use the crop recommendations model by supplying information about agricultural land.
a. Add Crop: contains details like your login, phone number, location, the type of crop you are growing, how many hectares you have, and the kind of seed you are planting.
b. View Crop Details: It displays all the details that are given to farmers so that they may comprehend plant production and availability in order to conquer crop production.
c. Crop Demand: We can obtain a forecast of the crop’s current demand by supplying the crop and location.
d. Crop Recommendation: This module, which is based on the knn algorithm, proposes the kind of plant we can plant depending on the soil type, pH value, and other characteristics.
For division (frequency) and regression, the Close Neighbor method, which belongs to the field of Supervised Reading, is employed. It is a versatile approach that may be used to resample data sets and insert missing values. K’s Nearest Neighbor, as the name would suggest, relies on K neighbours (Data Points) to forecast the class or continuous value of the new Data-point.
IV. RESULTS AND ANALYSIS
The android-based decision-making model was created to support farmers by giving them accurate and complete information on which crops are in demand and which should be planted on their land based on soil fertility.
The farmers provide their name, email address, mobile number, and password to register on the website. The user’s information is verified, and if it is accurate, the account is activated. If the information is false, an error notice is displayed. The user can then log in using their registered mobile phone and the password they chose when registering. Additionally, the database stores the user’s information for later use.
After successfully logging in, the user provides the relevant crop information, such as the crop they cultivate, the month they grew it, the number of acres they grew it on, and the number of seeds they planted in their agricultural property. The farmers’ current location is retrieved in order to prevent the user from providing incorrect information and to verify that the data they supply is accurate.
Here, farmers can access all the agricultural information that has been submitted by users, including the location, type of crop, month, number of seeds sown on the specific acres of land, and expected yield.
As part of our proposed work, we created an Android-based framer assistance application using Java, and we also analysed and forecasted crop demand as well as made crop recommendations. We did this by first creating a dataset using the information we had collected from various websites, blogs, and other sources We chose the knn algorithm as the best algorithm that fits the regression model for this work, as mentioned in the implementation and results section, after implementing the decision-making model, predicting, and analysing various algorithms. We used the elbow technique to analyse, assess, and perform clustering using the knneans algorithm. As a result, we obtained a best decision making model to assist the farmer overall. A. Future Scope The extensive adoption of cutting-edge management aids is not feasible in a situation where there are many geographically distributed, relatively small agricultural companies that are individually managed. Few farms have enough customers to warrant hiring operations research personnel, even on a parttime or consulting basis. A certain amount of decision modelling is done in an advisory context by some government extension services, but the adviser-tofarmer ratio is rarely favourable enough for the majority of farmers to benefit from such an approach. It is simple to understand why decision modelling is not used more frequently in these situations. Similarly, it is difficult to feel upbeat about the likelihood of any significant change in the future.
 ‘Agriculture Marketing Using Web and Mobile Based Technologies’, Abishek A.G., Easwari Engineering College, Chennai, India(2018).  ’ An Android Application for Finding Friends in Your Location’, School of Electronic and Computer Engineering(2016).  ’ Virtual Fruits Market – An Application for Farmer’, Kalyani Khodaskar Dept. of Computer Science and Engg. Yeshwantrao Chavan College of Engineering, Nagpur(2020).  ’E-Agriculture For Direct Marketing of Food Crops Using Chatbox’, Usha Kiruthika Dept of Computer Science and Engineering, School of Computing, SRM Institute of Science and Technology Kattankulathur(2020).  ’ Damaged Road Detection in Rural Areas for Improving Agricultural Marketing’, 1S. Leninisha Part time (Ph.D) Research Scholar in Anna University Asst. Prof. IT Dept, Easwari Engineering College, Chennai, TamilNadu, India(2020).  ’A Framework for Next Generation Agricultural Marketing System in Indian Context’, Swapnil Shrivastava and Supriya N Pal ,Big Data Analytics Group Centre for Development of Advanced Computing (CDAC) Bangalore(2019).  ’Predictive Analysis to Improve Crop Yield using a Neural Network Model’, International Conference on Advances in Computing, Communications and Informatics (ICACCI), S. Kulkarni, S. N. Mandal, G. S. Sharma, M. R. Mundada and Meeradevi,Bangalore(2018).  ’Crop Yield Prediction Using Data Analytics and Hybrid Approach’, Fourth International Conference on Computing Communication Control and Automation (ICCUBEA), S. V. Bhosale, R. A. Thombare, P. G. Dhemey and A. N. Chaudhari,India(2017).  ’Automated farming prediction’, T. Siddique, D. Barua, Z. Ferdous and A. Chakrabarty, Intelligent Systems Conference (IntelliSys), London(2017).  ’Crop Yield Prediction and Efficient use of Fertilizers’, International Conference on Communication and Signal Processing (ICCSP), Chennai, India, S. Bhanumathi, M. Vineeth and N. Rohit(2019).  ’Spiking Neural Networks for Crop Yield Estimation Based on Spatiotemporal Analysis of Image Time Series’ in IEEE Transactions on Geoscience and Remote Sensing, vol. 54, no. 11, pp. 6563-6573,P. Bose, N. K. Kasabov, L. Bruzzone and R. N. Hartono(2016).  ’Agricultural Year-Round Planning Model for Market-oriented Farms’ J. Chaiwongsai, N. Boonthep, Y. Miyanaga , T. Cheosuwa, B. Innawong •School of Information and Communication Technology, University of Phayao, Phayao, Thailand £Chitose Institute of Science and Technology(2019).  ‘Crop Shop – An application to maximize profit for farmers’, Niket Chauhan, M. Krishnakanth, G. Praneeth Kumar, Prerna Jotwani, Utkarsh Tandon, Abhishek Gosh, Nishant Garg, Santhi V. School of Computer Science and Engineering, VIT Vellore.  ‘Effecting of Mobile Applications in Control Farmers’ Communication Behavior in Iraq’, Majeed Hadi Saleh AL-Hamdani, Raad M. Ismail Al-Khazarji, Omer K. Jasim Mohammad Director of Computer Center University of Fallujah Fallujah, Iraq.  ’Crop Price Prediction System using Machine learning Algorithms’, Pandit Samuel B.Sahithi , T.Saheli , D.Ramanika , N.Anil Kumar(B.Tech , Department of Information Technology , ANITS , Sangivalasa ,Visakhapatnam)(2020).  ’AgroConsultant: Intelligent Crop Recommendation System Using Machine Learning Algorithms’, Zeel Doshi,Rashi Agrawal,Subhash Nadkarni,Prof. Neepa Shah Head of Information Technology Department,Dwarkadas J. Sanghvi College of Engineering Mumbai, India,email@example.com.  ’Crop Type Identification and Mapping Using Machine Learning Algorithms and Sentinel-2 Time Series Data’ ,Siwen Feng, Jianjun Zhao , Tingting Liu, Hongyan Zhang, Zhengxiang Zhang, and Xiaoyi Guo.  ’Comparative Analysis of Soil Properties to Predict Fertility and Crop Yield using Machine Learning Algorithm’,Pranay Malik ,Sushmita Sengupta,Jitendra Singh Jadon Department of Electronics and Telecommunication Amity School of Engineering and Technology Noida, India firstname.lastname@example.org(2021).  ’Predict Crop Production in India Using Machine Learning Technique: A Survey’Bhawana Sharma,Jay Kant Pratap Singh Yadav, Sunita Yadav, Department of Computer Science, Ajay Kumar Garg Engineering College, Ghaziabad email@example.com(2020).  ’Topdressing Nitrogen Demand Prediction in Rice Crop Using Machine Learning Systems’,by Miltiadis Iatrou 1,*ORCID,Christos Karydas 1ORCID,George Iatrou 1,Ioannis Pitsiorlas 1ORCID,Vassilis Aschonitis 2ORCID,Iason Raptis 1ORCID,Stelios Mpetas 1,Kostas Kravvas 3 andSpiros Mourelatos 1(2021).  ’Zeel Doshi; Subhash Nadkarni; Rashi Agrawal; Neepa Shah’,Head of Information Technology Department, Dwarkadas J. Sanghvi College of Engineering, Mumbai, India(2018).  ’Amenity Agriculture App using Random Forest Algorithm’, Valarmathi; N. Srija; S. Ananthi; Bhuvaneshwari P; Elakkiya T; Nishithraa N, Department of Information Technology, M. Kumarasamy College of Engineering, Thalavapalayam, Karur(2022)  https://www.educba.com/machine-learning-algorithms/
Copyright © 2022 Arpitha I M, A ThyagarajaMurthy. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.