TaroHacks Hackathon

• 

  Object Detection with Accuracy

Inspiration

We found out that California is a big deal when it comes to growing and selling tomatoes. But there's a challenge: tomatoes can start to go bad in about two weeks! And when you think about the time it takes to pick, package, and ship them, that's not a lot of wiggle room.

What it does

Our platform offers a valuable service to farmers by enabling them to submit images of their crops for analysis. Leveraging sophisticated machine learning technology, our system accurately assesses the current ripeness of the tomatoes depicted in these images. Moreover, by incorporating shipping logistics information provided by the farmers, our platform can forecast the expected freshness of the tomatoes upon reaching their intended destination. This optimization ensures that consumers receive the utmost freshness in their produce, aligning with our commitment to delivering the highest quality tomatoes available.

How we built it

In the development of our project, we leveraged a robust tool known as Yolov8, which functions as a valuable resource for enhancing computer image recognition capabilities. Yolov8 operates seamlessly within the Python programming language environment and is accompanied by an extensive image dataset. Our implementation involved enabling our computer's camera using Python and fine-tuning a scaled-down version of the Yolov8 dataset to facilitate the precise identification of ripe and unripe tomatoes. This revealed that training the computer for tomato recognition demanded substantial computational resources, particularly during graphic processing-intensive tasks.

Challenges we ran into

In our project, we ran into some big problems. First, we had trouble with Python libraries because they didn't match the correct version. This made errors pop up and stopped our code from working correctly. We learned that we needed to use the correct version of Python so everything worked together smoothly. Next, our model didn't always give accurate results, which made it hard to trust. It's super important for us to fix this to get results that we can rely on. Lastly, while we were training our model, Google Colab sometimes deleted our files. This was risky because we could lose important data and progress points. To fix this, we have to save our files safely so our training doesn't get interrupted. Even though we faced these challenges, we're determined to solve them. We want to make our project better and get great results.

Accomplishments that we're proud of

We are very proud of three big things we did:

Object Detection Skills: We became really good at spotting objects in pictures and videos. This is like being able to quickly find a friend in a big crowd.

Training With Our Own Data: Just like how we learn better with our own notes, we trained our computer program using our own special set of data, and it worked really well.

Creating a Cool Online Platform: We built a website using Next.js where people can use our object-spotting tool. This is like making a clubhouse where friends can come and see our cool projects.

These successes prove how good we are at computer stuff and how much we care about helping people all over the world.

What we learned

Here's a breakdown of some cool things we learned: Better Communication: We practiced talking clearly and listening to each other. This was super important because, without it, everyone would just jump in with their ideas and it would get confusing. Coding in Javascript: We learned how to code in a language called Javascript. It's like learning a new way to give instructions to a computer. Deep Learning in Python: We also got to know another coding language called Python and explored something called "deep learning". This is like teaching our computers to think and learn like humans. All these skills are going to be really helpful when we work on tougher projects in the future!

What's next for TaroHacks Hackathon

The Future of TaroHacks Hackathon Here's what's coming up for us at TaroHacks Hackathon: Joining More Hackathons: We're planning to be part of even more hackathons. It's like joining a sports tournament but for tech stuff, and it's super fun. Why We Love Hackathons: They are like big parties where we learn new things, come up with cool ideas, meet new friends, and grow as tech enthusiasts. So, for anyone interested in exploring the world of technology and innovation, hackathons are the place to be! We can't wait for what's next.

Built With

• github
• google-collab
• html
• javascript
• python