Diets require tremendous amounts of self control and determination. That's why every little thing which makes it easier to bear is a welcoming innovation. The more convenient something is, the easier it is to keep doing it. Therefore, the iHealthy aims to make the process of tracking your diet simpler.

What it does

The iHealthy is a convenient way for individuals to keep track of their caloric intake through the use of machine learning and object recognition. The user simply has to place their food item in front of the Jetson's camera. The Jetson will store information about the food item such as the name and nutritional information. Our product so far can identify name and number of calories for an item.

How we built it

Without a weighting device nor or any pressure sensors, we had to devise a method of detecting or identifying proportion (half a pizza does not equate the same as a whole pizza). We intended to gather images from various sources using our web crawler and Microsoft Bing of various foods of different rations. Use Digits to model them and BAM our product. But we ran out of time.

Challenges we ran into

We did not have an adequate amount of time to prepare. Consequently we had to spend the majority of our time at the hackathon brainstorming and learning how to use the tools given to us. Due to this we did not have enough time to implement all of the features we wanted to and was not able to polish our product to the degree we had wanted.

Accomplishments that we're proud of

Despite not having any prior knowledge of the tools we were using, we were able to implement iHealthy to a usable degree. We are proud that we were able to finish a working prototype by the end of our first hackathon.

What we learned

We learned that we should always have some prior planning before coming to a hackathon. Although this is our team's first hackathon, we believe we have learned quite a bit in building a product in a constrained time frame. We hope our experience in MakeMIT will allow us to have more success in the future.

What's next for iHealthy

The next logical step for iHealthy is to expand on the functionality of the system. For example, we had originally wanted to use the Microsoft Cognitive Science APIs to train our system to improve upon its object recognition capabilities. Additionally, we had planned to write a program to search the web with the recognized objects to parse and store appropriate information.

Built With

  • beautifulsoup4
  • c#
  • c++
  • digits
  • jetson
  • microsoft-cognitive-service-api
  • python
  • stl
Share this project: