The problem DigiFarm solves

Online Crop Auctions:
Recognizing the challenges farmers face in market access and fair pricing, we've introduced an online auction platform for crops. This digital marketplace connects farmers directly with diverse buyers, fostering fair competition and ensuring competitive prices. The streamlined auction system simplifies the selling process, reduces transaction times, and expands market reach for farmers.
here we help farmers where they can sell the crops across the world and get the fair prices

Disease Detection:
Complementing crop selection, our second machine learning model focuses on early disease detection in crops. By analyzing images of crops, the system identifies signs of potential diseases, allowing farmers to take proactive measures and minimize crop losses. This approach enhances overall crop health and resilience.

Online Crop Auctions:
Recognizing the challenges farmers face in market access and fair pricing, we've introduced an online auction platform for crops. This digital marketplace connects farmers directly with diverse buyers, fostering fair competition and ensuring competitive prices. The streamlined auction system simplifies the selling process, reduces transaction times, and expands market reach for farmers.

Challenges we ran into

1.CROSS-ORIGIN CONNECTIONS
requesting from the frontend(react) to the backend(node) has created lot's of error but hardwork was the key

2.WEB_SOCKET IMPLEMENTATION
making different rooms,joining multiple users into different auctions and also maintaining the live chat was taking so much efforts,time and logic. it caused lot error and overhead.but overcame it by just slightly modifying the way of passing bi-directional data

3.CROP-PREDICTION MODEL
training the model was a big challange for us like which algorithm to use. we have use bunch of algorithms analyse the accuracy also removing the overfitting we picked one of our best algorithm and trained the model with approx 99% accuracy.

4.DISEASE DETECTION MODEL
here the challage was working with the images and CNN.but we have the three pretrained models of three crops like tomato,potato and pepper.so we had load the model and fine-tuned it.while finetuning the models they took a lot time but we got introduced to collab's GPU and trained in less time

5.RAZORPAY PAYMENT IMPLEMENTATION
implementing the payment system according to our requirements especially into live bidding platform was challenging task.