Inspiration
Fruits can often be hard to grow, and often time, for a beginner, even common fruits can seem daunting and scary without the proper background information for it. I aim to make information about fruits, their harvesting time, and how to grow them available at the click of an image. This way, the user can identify what fruits they have, information about them, how to grow them, their harvesting uses, and a lot more. The possibilities are endless!
What it does
The user can start off by uploading a picture of the fruit they need more information about to a website. Then, the website identifies the fruit in the image using image recognition software, and web scrapes for information about the given fruit. Once information has been received, it displays it for the user, increasing their knowledge on the fruit.
How we built it
I started off by hosting my website on Flask. Flask allowed me to essentially create a local server to host the website on. To the website, users can upload files, and the POST request would be processed by Flask. Afterwards, the image the user had uploaded was sorted locally, and using the Google Vision API, was identified and sorted into the fruit it was. The, FruitVice, a fruit API, was used to confirm the fruit indeed was a fruit, and once this had been confirmed, Wikipedia was web scraped for information, which was returned to the user and displayed.
Challenges
Using image recognition software always has numerous challenges. Originally, I was going to self train software, such as openCV or CNN. However, that would require numerous data sets and valuable time, which given the deadline, I was not able to do. However, Google Vision API helped me overcome this challenge, with image recognition software with pre trained data. By using this, I did not have to create my own data to train, or even train my model at all, as Google had pre trained software available. Furthermore, another challenge was the POST request handled by Flask. This required various testing, debugging and more, as not only did the user need to send a POST request, internal handling of the image, and redirecting also had to be done.
Whats Next
This is an extremely useful solution to numerous people wanting to learn more about fruits they have, how to grow or harvest them, and general facts about them. I aim to make this more publicly available to everyone by hosting it on a public server, and also broaden the scope of information from just fruits to all foods in general, and even including things like recipes and more, making it applicable to everyone and useful to numerous people as well.
Log in or sign up for Devpost to join the conversation.