Efficient determination of disparity map from stereo images with modified sum of absolute differences (SAD) algorithm

Abstract

This paper proposes modification of the conventional Sum of Absolute Differences (SAD) for performance improvement in depth-map estimation from stereo images captured by a camera in a stereo system. The conventional SAD is commonly search in whole stereo images to find out the difference in pixels between the left and right captured images, and then obtains the corresponding disparity map and this may lead to high elapsing time. In order to reduce the number of searching pixels, the proposed modified SAD tries to estimate the difference only from edge pixels which are referred as pixels-ofinterest and bring significant information about depth map. The number of pixels being searched is reduced to about 17% on the total pixels, hence the total elapsing time is saved up to around 89% compared to that of the conventional SAD. This results is promising for implementation of a real-time vision system. Keywords-sum of absolute difference, stereo camera, disparity map, stereo visionI.

An FPGA stereo matching unit based on fuzzy logic

“…A general block diagram of the proposed method is shown in the Figure 1, the data obtained through a stereo camera is processed to the cadence of video and it is simultaneously computed three similarity parameters inherent to each pixel. 1correlation obtained via the Sum of absolute differences (SAD) algorithm [24][25][26][27]. 2 -similarity of the distance in pixels to the closer left edge.…”

Information Extraction and Object Tracking in Digital Video

“…This type of system has the function of detecting, classifying, tracking, analyzing, and interpreting the behavior of objects of interest. In [44,45], this technique was used combined with statistical techniques for controlling people's access to a specific location. It was also observed the use of intelligent monitoring systems, applied to building, port, or ship security [46,47].…”

A Review on Stereo Vision Algorithm: Challenges and Solutions

“…In 2010, Georgoulas, et al [2] implemented SAD in their proposed algorithm for real time stereo vision applications because of the low computational cost and simplicity characteristics. In 2013, it was improved to develop a modified version with better performance [38]. Posugade, et al [39] have tested the performance of the SAD method using a FPGA for real time application.…”

“…Limitations & Concerns approach SAD [34] Uses uniqueness Local CPU Good accuracy with low Not applicable for high of minimum and median computational cost texture images filter SAD [2] Uses pyramid reduction, Local FPGA Extremely low and Poor performance in ZNCC similarity measures consistent computational homogenous regions and vergence angle control cost and low power consumption SAD [21] Uses various sizes of Local CPU Gives good results in non-Performance evaluation correlation window texture regions as well as under radiometric depth discontinuity distortion is not justified SAD [38] Only considers object edge Local CPU Faster than conventional Performance evaluation pixels in depth estimation SAD algorithm under radiometric calculation distortion is not justified SAD/SSD [37] Uses nuclear norm Local CPU Able to reduce occlusion Performance in terms of minimization error and performs better in computational cost is homogenous regions unclear SAD [40] Uses bilateral filter Local CPU Good accuracy and is able to Performance in terms of reduce noise computational cost is unclear NCC [83] Modified to be more like Local CPU Faster than conventional Quality of disparity map is template approximation NCC not verified NCC [84] Uses shape adaptive Local CPU Produces accurate disparity Not suitable for real time window and orthogonal map quickly application yet integral image technique NCC [44] Uses dirty filtering Local CPU High precision High computational cost ZNCC [85] Integrated within neural Local CPU Good in dealing with Ineffective in dealing network textureless areas occlusion and discontinuity CT [54] None Local ASIC Low computational cost Disparity map quality is not justified CT [53] None Local FPGA Low computational cost Post-processing is not included, so disparity quality can be further improved improved CT [55] Optimizes size and Local FPGA High speed with reduced Poor performance when memory access memory size image has noise CT [86] Combines with SAD and Local CPU Fast operation with Computational cost is not uses permeability filtering promising accuracy low enough for real time application CT [50] Uses coherency sensitive Local CPU High accuracy with fewer High computational cost hashing bad pixels CT [58] Uses random walk and Local CPU Minimizes noise Parameter selection is wavelet edge joint bilateral significantly, even in the crucial to ensure optimum toughest situation with performance addictive Gaussian noise DP [64] Uses generalized ground Global CPU Scanline inconsistency Not applicable for real control points scheme problem is removed time application DP [87] Uses adaptive aggregation Global CPU Faster and better than Implementation on CPU conventional DP based has high computational approach cost DP…”

“…A lot of studies are done on reducing computational complexity of stereo vision algorithms. For software based improvement, Hoa, et al [38] reduced the number of pixels being searched of the SAD algorithm to only 17% of the total pixels, which lowers computational cost approximately 89% compared to the conventional SAD algorithm. In this study, only edge or corner pixels are selected for matching cost and disparity estimation while the rest are excluded from estimation.…”