Introduction to Image Processing
01.10.2017
This course focused on basic algorithms of digital image processing. In a group project our task was to develop a more or less useful tool using MATLAB and programming two of the learnt algorithms from scratch.
The methods included:
- Hough Transformation
- Mathematical Morphology (Dilation, Erosion, Skeletonization)
- Local Binary Patterns
- Histogram of Oriented Gradients
- Integral Images
- Image Pyramids
- Region Growing
- Connected Component Labeling
- Geometric Transformations
- Optical Flow
- Distance Transformations
- Canny Filter
- Otsu Threshold
- Template Matching
- Sobel Filter
- Prewitt Filter
- Roberts Filter
- Laplace Filter
- Gauss Filter
- Projections
Our project was called the Puppetizer. The goal of this tool, was to extract a person from an image and place them in front of an arbitrary background in an arbitrary position. By using the MATLAB integrated threshold function, we extract the person from the background. We skeletonize the body, identifying the body parts and with geometric transformations we position the single parts in a pre-determined position and a chosen background. With this project we were awarded the best of the course.