Computer Vision: What Engineering Field to Study?

AI Thread Summary
Computer vision is an interdisciplinary field that combines computer science, applied mathematics, and statistics, focusing on pattern recognition and data mining. Those interested in computer vision should consider their specific interests: if they lean towards image processing algorithms, a specialization in computer science is advisable. Conversely, if the focus is on hardware development for computer vision applications, electrical or computer engineering may be more suitable. The field encompasses various specializations, and collaboration among electrical engineers, software engineers, roboticists, and physicists is common. Ultimately, students should align their interests with their university's course offerings, exploring areas like opto-electronics, photonics, and optical engineering, especially if they are interested in systems for tracking moving targets, which require a blend of math, physics, and engineering disciplines.
tbarker5
Messages
10
Reaction score
0
I'm first year ENG at a Canadian university where first year is general for all engineers. Next year I get to pick a field to specialize in. I'm really interested in computer vision, for example a system that could track a moving target or something. I Know this is crazy complicated but I'm wondering what field would be best to go into if this is my biggest interest. I'm not sure if it would be a purely software engineering problem, mechatronics, electrical, computer or some combination of multiple. If it is a combination of multiple which do you think would be the best engineering stream to go into if this is my primary interest?
Thanks in advanced for the feedback
 
Physics news on Phys.org


Hey tbarker5.

Computer vision is a combination of theoretical (and practical) computer science with applied mathematics and statistics.

You are looking at the area of pattern recognition and this is a huge field with a lot of specializations (and further specializations within those) that cover everything from finger-print and other biometric data recognition and classification to the more general techniques which are studied under the field known as data mining.

You also have fields that study the same sorts of things but are under a completely different name and this is one reason why you get people in different areas you discover the same things but are unaware simply because of the disconnect that different fields have with one another under hyper-specialization.
 


I've taken computer vision-related courses in the graduate level, and I also know a few computer vision researchers who do industry research at Microsoft Research.

Computer vision (or CV as it's often called) is indeed very interdisciplinary, and you could easily find electrical engineers, software engineers, roboticists and physicists who specialize in optics working together on computer vision research or development.

You have to ask yourself what part of computer vision interests you the most: is it the image processing algorithms that take raw pixel data and give it semantic meaning with regards to the program? If so, you're probably going to want to specialize in computer science, especially if your program has undergraduate courses available in the subject.

If, instead, you're interested in building hardware that makes it possible to apply computer vision algorithms (like building the next Microsoft Kinect, to give an example), then you might want to look into electrical engineering or computer engineering. It's quite possible that you would do vision-related work in EE courses (I took a digital systems course where we did image processing with FPGAs that we programmed in Verilog).

Ultimately, you're going to have to find the intersection between your school's course offerings and your own interests. But if I had to make an uneducated guess I would say that computer science/software engineering is where you'll have the most success pursuing vision-related work.
 


Maybe not entirely what you're looking for, but have you considered opto-electronics, photonics, or optical engineering? If you're looking for systems that can track a moving target then look no further than modern avionics, which is a ton of math, physics, electrical engineering, optics, materials, signal processing, computer programming, etc. Check it out.
 
I graduated with a BSc in Physics in 2020. Since there were limited opportunities in my country (mostly teaching), I decided to improve my programming skills and began working in IT, first as a software engineer and later as a quality assurance engineer, where I’ve now spent about 3 years. While this career path has provided financial stability, I’ve realized that my excitement and passion aren’t really there, unlike what I felt when studying or doing research in physics. Working in IT...
Hi everyone! I'm a senior majoring in physics, math, and music, and I'm currently in the process applying for theoretical and computational biophysics (primarily thru physics departments) Ph.D. programs. I have a 4.0 from a basically unknown school in the American South, two REUs (T50 and T25) in computational biophysics and two semesters of research in optics (one purely experimental, one comp/exp) at my home institution (since there aren't any biophysics profs at my school), but no...
Back
Top