Classical Vision Techniques

Jan 8th, 2022

Recognising interesting points such as corners, edges and/or blobs in an image, identifying features among these interesting points that are both scale-invariant and rotation-invariant, extracting such features and matching them with similar features from other images is essential for most of the robotic applications. This article explores some of the underlying fundamental principles of classical state-of-the art feature extraction techniques.

Feb 5th, 2022

As roboticist, we understand that not all the pixels in an image are informative in robotics, especially for applications such as grasping, object detection and/or reconstruction. This article explores some of the rudimentary processing on an image with an example algorithm in MATLAB.

Feb 7th, 2022

While image processing is particularly important in robotic vision, most of these techniques are also adapted as pre-processing steps or encoding schemes in machine learning approaches. This article explores some of these all-important techniques with example algorithms in Python.

%d bloggers like this: