InvoPedia

Conservation efforts often overlook civilian potential in the eradication of invasive plants. InvoPedia grants everyone the power to support and protect their local ecosystem.

Comment

Inspiration

Living in Canada, we are blessed with vast expanses of pristine, biodiverse wilderness. However, the growing threat of invasive species is putting our natural ecosystems in jeopardy, and it can be easy to feel powerless in the face of this challenge due to the fact that there is a lack of information on how we as ordinary citizens can help. With InvoPedia, we hope to make information on invasive species more accessible, and to enable everyone to exterminate invasive species in their local environment and help conservation efforts.

What it does

InvoPedia is a web-based application that allows users to upload or take a picture of a plant and identify if it's an invasive species, what it is, and how to exterminate it. For example, if a user sees a foreign plant(e.g. purple loosestrife) and uploads a picture of it to InvoPedia, it will identify that it is an invasive species, what species it is specifically, provide information and recommend actions that can be taken to eradicate it.

How we built it

We created a computer-vision model using Microsoft Azure's custom vision that identifies various invasive species. Then, we created a web application with the help of Microsoft Azure's app services. The web application called the API of the model in order to give its prediction, and used the cohere API to create natural language generation for recommendations & information. In addition, there is an option to report the invasive species to the Toronto Region Conservation Authority. We used JavaScript HTML and CSS to create/style the website and call APIS.

Challenges we ran into

  1. Training the computer vision model
  2. We found that there was a lack of data resulting in an inaccurate computer vision model at first 2.Determining the forms of deployment of model
  3. Issues of compatibility of various aspects of our project arose while putting the project together and making it a functional project 3.Working with API’s and javascript
  4. One of our biggest challenges was finding the compatibility and solution to API connections resulting in difficulties to link them together
  5. We had used HTML and Javascript for the first time and were facing many difficulties as is found when working with a new language ## Accomplishments that we're proud of The past 2 days have been a journey with many ups and downs. With any project, one will have accomplishments and difficulties. For us, we felt accomplished while slowly progressing through our challenges. One such instance is once we were able to integrate the front-end to the back-end with the help of Microsoft Azure. The Microsoft products have been a great help to us from the brainstorming phase to the final product and without it we wouldn’t have had such a successful project. Another instance was while connecting the hardware to software allowing the user to utilise a mobile version of our product. We found many compatibility issues and issues connecting the device to the program. Overall our greatest accomplishment was our ability to organize ourselves while keeping communication channels open and reaching to agreements on many aspects of the project allowing us to complete the project effectively and efficiently ## What we learned Throughout the progression of the project, we were able to effectively learn many new skills and topics. One such instance was while using and learning Javascript and HTML. While building our project, we gained valuable knowledge and experience in working with Javascript and HTML, making us feel more comfortable programming in these languages. Another thing we learned was how to train a computer vision model and integrate it into our website. We used Microsoft Azure to train the model which clearly explained how to train the model but our issue was finding decent data. We overcame this and learned that finding data is through a massive amount of googling. Overall the project was a success in our view and we we able to learn many new skills and topics. ## What's next for InvoPedia
  6. We would like to transition from a cloud-based web app to a mobile application in order to create a more customizable a localized experience for users
  7. Creating location-specific algorithms for invasive species recognition to be more efficient and accurate
  8. Adding options to report incidents to local conservation authorities using location data

Built With

Share this project:

Updates