📄️ Overview
Images are pixels on a 2D grid. This is a type of densely connected graph, where each pixel in a quantized local region is a node and each neighbor is an edge. We can break images into regions using KMeans or SLIC. In this case we want to track neighboring relations between these regions. We convert images into a Graph datastructure consisting of Node nodes. Each Node is a collection of pixels representing a unique region. Each Node tracks its immediate neighbors. The Graph manages all Nodes.
📄️ Graph and Node Explained
To explain Graph creation from image regions we show a diagram representation followed by a concrete step-by-step example.
📄️ Important Functions Explained
Nodes are responsible in computing their own contours using the Suzuki-Abe method.
📄️ API / Usage
Each Node is a collection of pixels. A Node holds a unique_ptr to a vector of pixels with RGBXY structure. \