I am pursuing my Ph.D. in the wireless sensor network and hopefully, I will complete it in about a year. But in my next career, I want to do my research on Computer Vision. Can anyone guide me to go ahead with fundamental things to do research in this field?
I see an increase in autonomous cars. While most people think this is futuristic most cars aid the driver using a computer of some sort like antilock brakes. How about a system that both identify the driver and impaired by drug consumption or fatigue. This dual authentication and authorization process could potentially be bought by parents and they could be notified if their teen is too impaired to drive home.
HI
Computer vision is concerned with the automatic extraction, analysis, and understanding of useful information from a single image or a sequence of images.So to start with analysis of image will be better.
Hai Sir
I would prefer you to do your research in the area of audio-visual speech recognition or gesture recognition which targets the physically challenged people.
Dear sir, I would recommend you to learn some basic digital images processing. The basic understanding of digital images processing will help you to grow faster in computer vision path. https://www.amazon.com/Digital-Image-Processing-Rafael-Gonzalez/dp/0133356728
Finding a problem that needs visual computing (image based /video based)
I agree with Aparna Sathya Murthy. Nothing motivates you more than having a real problem to solve. You learn the stuff that the problem demands, you learn about the various approaches to particular problems, and if you are smart, you learn the underlying theory as well so that you understand how and why the various approaches fail.
I also agree with the people who suggested that you start with image processing, but always keep in mind that image processing allows you to transform images so that people can detect, localize and quantify the features of interest whereas computer vision attempts to use the computer to do the detection, localization and quantification of them. Thus, whole areas of image processing really don't apply to computer vision. After all, if the compute can discern the difference between two areas of an image so as to enhance that difference for a human observer (or "receiver" in the parlance of signal communications), then the computer might as well just make the required decisions.
I taught computer vision for about 18 years.
Best regards,
Peter
Dear Sir, Computer vision is basically a image processing field using software platform. It contains hell lot of algorithms.Currently, Software platform may be MATLAB, OpenCV Library, LABVIEW, GeoTIFF library(For remote sensing application eg Images taken by satellite) etc.I have worked on one project namely Computer aided system for paralytic patient.And computer vision helps me a lot to complete my project work. For computer vision to understand very well,one should be very good in mathematics.Because if we see the algorithms specially implemented for CV are nothing but maths (Most appropriately it is statistics,numerical methods and probability theory). So,If one has to succeed in CV, he/she has to go through mathematical models. Following are the simple steps one can follow: 1. Chose the research topic 2. Find some already available material for that topic( If earlier work is there if any) 3. Which are the difficulties they have faced and what is the future scope ? What we can add / modify /optimize ? 4. Download some papers on the particular topic. 5. Choose one paper that you understood at least 30-40 % after just reading. 6.Now try to find out the mathematical terms author have mentioned 7. Google those mathematical algorithms and try t get that.. 8. And then hit the software
You will definitely get the success.... All de best.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing