OpenCV_Work
OpenCV Work, including interesting problem statements [following OpenCV Tutorials - Python]
Digital Image Processing is the step-1 to start with Computer Vision. Think of a problem of scene recognition, what is the first thing that comes in mind after seeing a scene? Several objects, background, some moving and some static.
A problem can be subdivided into sub-problems, and sub-problems into their sub-problems. In this way, if carried on, the basic problems come down to - Processing the images according to our need. What is our need in detecting features? A grayscaled image? Thresholded? Mean blurred? Edge detected? There are many more ideas to this, and this has been covered up in this project.
Two goals:
- Object Recognition / Scene Understanding.
- Template Matching.
How the goals have been targetted? Both of the goals are nothing but complex computer vision problems, active research topics, but they are nothing more than simple problems for Humans. I've tried to think about them as a problem for my eyes, and how will I tend to solve this? For that, the filters I thought will be required, for the completion of goals, have been implemented - tested - compared. In this way, getting better knowledge of the filters and pre-processing of images.
Files:
-
blurring_.cpp, difference_images.cpp, screenshot_cpp.cpp: Includes blurring methods, detailed images, and screenshot taking code respectively.
-
cartoonification.cpp, draw_rectangle.cpp These are long programs, of cartoonification of an image and choosing ROI and applying filters & methods on it, respectively.
Cartoonification Project Report has been uploaded and can be checked for further details.
Suggestions, feedback, commits - welcome!