Computer Vision

What is Computer Vision?

Computer vision is a sub-set of artificial intelligence that tries to gives computers a visual understanding of the world. The primary goal is to emulate human level vision and understanding of the world around them. Computer vision has three main steps:

1. Image acquisition

2. Image processing

3. Image analysis

Image acquisition is relatively simple. It's essentially taking a picture with a camera, and putting it on your computer. The computer translates that picture into a series of ones and zeros. A digital image which is then ready for step three.

Image processing is the low-level analysis of images. It turns all those numbers into usable information characterized by image edges, point features or segments.

Edge Detection

Image Segmentation

The third and final step analyzes this information and makes decisions. By looking at the low-level information and comparing that to what it already knows, the computer can infer certain higher level facts. This could be the classification of different objects in the image (or video), object tracking, or facial recognition. The technology is used in self driving cars, security cameras, and robots like pepper.

Our project is specifically designed to analyze images of food and tell us their nutritional qualities. Our algorithm can determine if food is vegetarian, vegan, gluten-free, a nut-allergen, and the amount of calories per serving. This can be very useful to people with allergies and dietary restrictions. See Application for more information.

Google Sites

Report abuse