I wrote an image processing algorithm on MATLAB and I want to know the CPU time consumption of the same algorithm if it was written using OpenCV.
The only way to answer your question is to write the same algorithm using OpenCV and time it on the same machine.
Dear Ahmed,
if you give more details you can have a better answer.
What algorithm are you interested in?
What is the size of your images?
etc...
The only way to answer your question is to write the same algorithm using OpenCV and time it on the same machine.
@Dima: You are right. Also, on the same machine it depends how opencv are compiled (e.g if you are using tbb or not), how well your code is parallelized, how well your opencv code is written, how well the matlab code is written, etc...
However, you can get a general idea. In my past experiments, I found that on average opencv is 40 - 50 times faster than matlab. This depends on the functions you are evaluating, of course.
Dear Michele
I am working on license plate detection algorithms. I used learning methods like LDA, morphological function also
image size is 640 x 480 i want to know time consumption of my algorithm if it was written in OpenCV
thanks for help
I recently tried this and the comparison is actually algorithm dependent. My algorithm for path detection which took about 500 ms on Matlab is taking around 120 ms on OpenCV. Simple morphological operations tried out on Matlab are much faster (around 40-50 times) as mentioned above by Michele. So, I'd suggest that you port it to OpenCV and check it.
hi Ahmed,
firstly Its algorithm dependent,
secondly are you processing every pixel or how u detect your ROI (region of interest) are to be consider things.
plz go through the following link it may give some basic things on LDA
http://papers.nips.cc/paper/2547-two-dimensional-linear-discriminant-analysis.pdf
CPU time consumption is only dependent on your coding as well as your algorithm compilation process
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering