During my time at Niagara Bottling, I made a Computer Vision Model that was trained using machine learning on a dataset that I personally annotated

The software can detect a pallet of water bottles, the cases that make it up, and all the relevant data, such as labels, brands, sizes, etc.

This is a more sophisticated model that doesn't rely on boxes and can detect the shapes of the individual cases alongside the overall pallet of bottles.

Optimized Python code to add the functionality of detecting QR codes, barcodes, and label text to the OAK Camera hardware in the plants. This is then compared between the separate sources to make sure that the data was correctly observed.

Utilized the Google Cloud Vision API to find ingredients from a video stream and match them to a recipe in an XML file.

Developed a website using HTML and CSS that takes the aforementioned information and displays it to an end user.

Designed a user interface that effectively communicates how much of each ingredient needs to be added at any moment, and displays a pop-up whenever an excess amount or an unnecessary ingredient is detected.