Contains laboratory, test and home works for Image Processing classes of Polevoi Dmitry @ NUST MIS&S Details:
- Laboratory work 1: Creating binary mask, using thresholds and applying it the the original image
- Laboratory work 2: Auto-contrasting using histogram analysis and applying linear-transform via lookup table
- Laboratory work 3: 2D-Convolution using OpenCV
- Test work: Generating image, finding different gradients of it, calculating module of the gradient and combining them into the new image
- Homework: Creating the maps of heights of images and training U-net network to segment the source images using them
- Homework pt.1: Preprocessing dataset - Creating the maps of heights
- Homework pt.2: Training network - Creating U-Net in MxNet Framework and training this model to get segmentation of RGB images from CityScapes