Image Processing and feature detection are the two most important areas of research in Computer Vision. In the one year project we successfully implemented various techniques and coding schemes that deal with Image Processing and Feature Detection of both still images and live video under the mentorship of Dr Amita Kapoor.
We started with converting raw images to usable forms like grayscale, plots. We performed various mathematical operations on the image transforms like blurring (Gaussian Transform), sharpening etc. We learned to perform weighted addition of images. On both still images and live video features like face, eyes, smile, upper/lower body and full body have been detected. In this project, we used Haar Classifier to detect human face and eyes. We also employed the use of support vector machines to successfully identify human upper/lower and full body. The work can be extended to pedestrian detection in case of self-driving cars.
amita-kapoor/Image-Processing-and-Feature-Detection
Image Processing and feature detection are the two most important areas of research in Computer Vision. In our one year project we successfully implemented various techniques and coding schemes that deal with Image Processing and Feature Detection of both still images and live video. We started with converting raw images to usable forms like grayscale, plots. We performed various mathematical operations on the image transforms like blurring (Gaussian Transform), sharpening etc. We learned to perform weighted addition of images. On both still images and live video features like face, eyes, smile, upper/lower body and full body have been detected. In this project, we used Haar Classifier to detect human face and eyes. We also employed the use of support vector machines to successfully identify human upper/lower and full body. The work can be extended to pedestrian detection in case of self-driving cars.
PythonApache-2.0