/Probability-Based-Boundary-Detection

This Repository consists of implementation of probability based boundary detection algorithm which gives more accurate results than canny edge detection.

Primary LanguagePython

Probability-based-Boundary-Detection

This repository consists of comparison of baseline edge detection algorithms like Canny and Sobel with Probability of boundary detection algorithm. A simpler version of PB algorithm has been implemented which considers texture, color and intensity discontinuities. This algorithm predicts per pixel probability of the boundary detected. The original image and the output of implementation is shown below:

Original PBLite

The algorithm of PBLite detection is shown below:

PBLite

The main steps for implementing the same are:

Step 1: Feature extraction using Filtering

The filter banks implemented for low-level feature extraction are Oriented Derivative if Gaussian Filters, Leung-Malik Filters (multi-scale) and Gabor Filter.

DoG PBLite PBLite

Step 2: Extracting texture, color and brightness using clustering

Filter banks can be used for extraction of texture properties but here all the three filter banks are combined which results into vector of filter responses. As filter response vectors are generated, they are clustered together using k-means clustering. For Texton Maps k = 64 is used; Color and Brightness Maps k= 16 is used.

DoG PBLite PBLite

The gradient measurement is performed to know how much all features distribution is changing at a given pixel. For this purpose, half-disc masks are used.

PBLite PBLite PBLite

Step 3: Pb-Score

The gradient maps which are generated are combined with classical edge detectors like Canny and Sobel baselines for weighted average.

output

Run Instructions

python Wrapper.py

File structure

├── Code
|  ├── Wrapper.py
|  ├── utils.py
├── BSDS500
├── Results
|  ├── color_gradient_maps
|  ├── color_maps
|  ├── intensity_gradient_maps
|  ├── intensiy_maps
|  ├── final_output
|  ├── texture_maps
|  ├── texture_gradient_maps