I am posting an outline on research relating to computer vision. A high conceptual model of the field is helpful when exploring. Below is just a short descriptions with examples.
- Low Level Vision
- most basic of image analysis.
- This level can compare images with basic image processing and optical flow.
- hashing algorithms in comparing images
- ex https://github.com/moondra2017/Computer-Vision
- Mid Level Vision
- basic segmentation and fitting is the focus for features extracted at this level.
- Useful for identifying shapes and lines in an image.
- Multiple View Vision
- combine images from different view for a more comprehensive vision.
- Highly useful for self driving cars and 3d scenes.
- ex. https://github.com/alicevision/meshroom
- High Level Vision.
- object detection and classification is the popular sought out tasks to understand.
- Deep neural networks are also on this level.