https://www.youtube.com/watch?v=UPafVnSR1dY&embeds_referring_euri=https://devpost.com/&feature=emb_imp_woyt
https://github.com/Srikark-17/Farmerce
With covid-19 restrictions being lifted and local farmer's markets starting up in many places, people have a chance to expose themselves while being outside. Farmerce aims to reduce the chance for exposure while acting as a platform that connects customers to vendors by showing them information such as the methods of production, contact information, and items for sale. This will revolutionize the agricultural industry through an easy-to-use web app that gives users important information.
Farmerce is an online platform that connects customers to vendors at local farmers markets. It provides customers the option to view information about a specific vendor such as items for sale, contact information, and methods of growing. It also provides users the chance to take pictures and upload it to see if has a disease or not. It also shows local farmers markets based on the region.
Our target audience for farmerce is the agricultural community(shoppers and farmers). Shoppers will access the web app in the car, or at their homes. They will then proceed to view information about the farmers, items for sale, and learn more about the produce. When they are in the car, they can use our "find farmers markets near me" feature to find the closest market. After arriving at that market, shoppers can go around and look at stalls. If a piece of produce looks rotten or black, they can use our plant disease detection AI to determine if it is safe to buy. Farmerce allows shoppers to save time, have less exposure to covid, and make sure that they are buying healthy vegetables.
We utilized flask framework to integrate our expertise in python in the creation of a web application. With this we were able to create a Convolutional Neural Network that helps farmers identify crops, use MongoDB to dynamically post and view farmers at a market place: their reputation, ratings, contact, chemicals involved and more, find local farmers market near you and ultimately create a networking platform to provide more information and build trust between the producer and the consumer.
We used mongoDB to dynamical retrieve data from our forms to create the posts We used mongoDB to dynamically post data in realtime to our database later to fetch We used a CNN (Convolutional Neural Network) to help farmers identify bad crops We used a CNN to classify the images of items being post to help verify that products being sold through our platform are indeed healthy We used Keras and sklearn to effectively create a CNN that identifies plants with diseases. We obtained a 97% accuracy by training our data on a separate machine for 3+ hours.
Implementing Mongo DB to create a unique profile for each farmer Our UI design, which makes our app user-friendly. Creating an AI to detect plant diseases.
Many of us were not very experienced ins the field of web development or python, and those who were were not experts at both. This project allowed us to explore and expand our expertise into unfamiliar territories such as integrating flask and using a new database system such as MongoDB.