Real-World Applications of Computer Vision

From health applications to the autonomous vehicle space to sign language translation for the deaf and hard of hearing, computer vision skills are needed in a variety of fields. Here are just a few examples:

Hand Recognition

Before learning how to train machines to identify objects in space in real time, it’s important to understand how still images are analyzed and recognized. For example, in Lesson 4: Types of Features and Image Segmentation, you’ll learn about a technique called contouring, which is used in hand recognition programs. Hand recognition is a widely-recognized way of facilitating HCI (human-computer interaction), which is used in a variety of ways from mobile applications to gestural interpretation, so the foundational knowledge of not just how to program it, but how it works, is critical for those who may be looking to pursue a career in computer vision.

Then, you can take these skills and eventually teach a computer algorithm to track hand gestures in real time. A quick search on Youtube will turn up many interesting examples of how hand recognition can help with language processing, augmented reality, and even the creation of abstract art!

Self-Driving Cars and Spatially Coherent Data

Computer vision is used for vehicle and pedestrian recognition and tracking (to determine their speed and predict movement). Check out this blog from David Silver on how computer vision works for self-driving cars.

Medical Image Analysis and Diagnosis

In addition to founding Udacity and starting the Google X and Self-Driving Car labs at Google, Sebastian Thrun is also a professor of computer science and an extensively published artificial intelligence scientist. In early 2017, his team at Stanford produced a report on how deep learning--which is tied intrinsically with computer vision--can be used to identify different types of skin cancers at an accuracy level comparable to that of dermatologists.

In the Extracurricular section of your classroom, you’ll find a course taught by Sebastian himself that takes you through the research his team did to identify melanoma lesions, and what implications this research has for the future of medicine. The course material should take you between two and three hours, and at the conclusion, you’ll have the opportunity to write an implementation of your own.

While you technically don’t need to take the course to graduate, understanding and being able to discuss applications of computer vision in the real world can really impress a hiring manager during your interview or technical interview. It can also open the door to job opportunities in tech-adjacent fields--in this case, medical technology.

Up Next

By now, you have made progress in your Nanodegree program and gotten an overview of jobs in computer vision. You should feel comfortable seeking out information online to further your knowledge of the community and new developments in the field. You should also know how to find computer vision and related jobs on job boards.

Continue on in the lesson to learn how to leverage Udacity Career Services in your career development.