Image Segmentation and Feature Extraction Using Sift-Sad Algorithm for Disparity Map Generation

Abstract

In this paper, a stereo matching algorithm based on image segments and disparity measurement using stereo images is presented. We propose the hybrid segmentation algorithm that is based on a combination of the Belief Propagation and Mean Shift algorithms with aim to refine the final disparity map by using a stereo pair of images. Firstly, a color based segmentation method is applied for segmenting the left image of the input stereo pair (reference image) into regions. The aim of the segmentation is to simplify representation of the image into the form that is easier to analyze and is able to locate objects in images. Secondly, results of the segmentation are used as an input of the SIFTSAD matching method to determine the disparity estimate of each image pixel. This matching algorithm is proposed by combining Scale Invariant Feature Transform (SIFT) with the Sum of Absolute Difference (SAD). Finally, the comparisons between the three robust feature detection methods SIFT, Affine SIFT (ASIFT) and Speeded Up Robust Features (SURF) are presented. The obtained experimental results demonstrate that the proposed method has a positive effect on overall estimation of disparity map and outperforms other examined methods.