I am currently working with computer vision with an ultimate goal of building a sign language trainslator
NB: I am using better comment extension, this might make some of my comments seem wired to you
The learing process is really interesting, these are what I have accomplish:
- A blue Screen minipulator script screen that removes the blue screen from an image and replaces it with another background
- A green screen minipulator script similar to that of the blue screen
- A colour selection script (currently for pink)
- A day and Night image classifier that uses the H and V channel to classify the image, currently at 97.5% accuracy
- Created an edge detection script that implemented fourier transform
- Created a high pass filter for detecting edges. The filter takes in a Karnel and a gray image and then returns a binary edged image
- Implemented low pass filter
- Implemented Canny Edge Detection
- Implenented shape detection using Hough transformation
- Built a face detector using Haar Cascades
- Studies feature detection and wrote some sample codes for contouring, corner detection and segmentation using K-mean
- Learnt about the ORB algorithm and how to implement OpenCV's version of the algo for image recognition
- Learnt about HOG algorithm
- Learn about CNN and implemented a simple CNN using PyTorch
- Exploring the FashionMNIST Data Set
- Explored Single Shot Detector (SSD) and MobileNets for object detection