Method

Appearance-Aligned Block-Matching Stereo [AABM]


Submitted on 17 Sep. 2015 11:42 by
Nils Einecke (Honda Research Institute Europe GmbH)

Running time:0.08 s
Environment:1 core @ 3.0 Ghz (C/C++)

Method Description:
You can get a copy of the paper here.

Appearance-aligned stereo warps one stereo image according to the expected road geometry for reducing the perspective appearance change between the left and right stereo image. The resulting disparity map is combined with the disparity map of stereo computation without warped images by means of a pixel-wise correlation value comparison. This way warping model coherent pixels are mainly drawn from the aligned disparity map and model incoherent pixels are drawn from the normal disparity map.

For the actual stereo computation a standard block-matching stereo with SNCC cost function is used. The images are pre-processed with a Gaussian filter and resized to half the image size. After the stereo processing the disparity maps are rescaled to full image size. Furthermore, standard post-processed is applied, including a left-right check, error island removal (region grouping), hole-filling and median filtering.

Timings: block-matching SNCC stereo (38ms), LR-check (18ms), region growing (2ms), hole-filling (2ms), median filter (23ms)

Runtime includes some IPP routines but no multi-core or GPU usage.
Parameters:
linear warping model: x2 = x + a*y + n
a=0.25, n=-21.6
SNCC (first stage 3x3, second stage 9x9)
sigma_gauss = 1.0
disp_range = 64
min correlation threshold = 0.3
region growing threshold = 2.5 disparity
min region size = 200 pixel
median filter = 1x9 and 9x1
Latex Bibtex:
@InProceedings{Einecke2013IV,
author = {N. Einecke and J. Eggert},
title = {Stereo Image Warping for Improved Depth Estimation of Road Surfaces},
booktitle = {IV},
pages = {89-194},
year = 2013
}

Detailed Results

This page provides detailed results for the method(s) selected. For the first 20 test images, the percentage of erroneous pixels is depicted in the table. We use the error metric described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), which considers a pixel to be correctly estimated if the disparity or flow end-point error is <3px or <5% (for scene flow this criterion needs to be fulfilled for both disparity maps and the flow map). Underneath, the left input image, the estimated results and the error maps are shown (for disp_0/disp_1/flow/scene_flow, respectively). The error map uses the log-color scale described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), depicting correct estimates (<3px or <5% error) in blue and wrong estimates in red color tones. Dark regions in the error images denote the occluded pixels which fall outside the image boundaries. The false color maps of the results are scaled to the largest ground truth disparity values / flow magnitudes.

Test Set Average

Error D1-bg D1-fg D1-all
All / All 4.88 16.07 6.74
All / Est 4.88 16.07 6.74
Noc / All 4.49 15.22 6.26
Noc / Est 4.49 15.22 6.26
This table as LaTeX

Test Image 0

Error D1-bg D1-fg D1-all
All / All 5.45 4.05 5.25
All / Est 5.45 4.05 5.25
Noc / All 5.42 4.05 5.23
Noc / Est 5.42 4.05 5.23
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 1

Error D1-bg D1-fg D1-all
All / All 3.71 11.60 4.59
All / Est 3.71 11.60 4.59
Noc / All 3.60 11.60 4.51
Noc / Est 3.60 11.60 4.51
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 2

Error D1-bg D1-fg D1-all
All / All 5.48 15.70 5.97
All / Est 5.48 15.70 5.97
Noc / All 4.92 15.70 5.46
Noc / Est 4.92 15.70 5.46
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 3

Error D1-bg D1-fg D1-all
All / All 7.57 19.43 8.67
All / Est 7.57 19.43 8.67
Noc / All 7.09 19.43 8.25
Noc / Est 7.09 19.43 8.25
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 4

Error D1-bg D1-fg D1-all
All / All 8.65 16.63 9.97
All / Est 8.65 16.63 9.97
Noc / All 7.91 16.63 9.38
Noc / Est 7.91 16.63 9.38
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 5

Error D1-bg D1-fg D1-all
All / All 10.74 5.26 10.25
All / Est 10.74 5.26 10.25
Noc / All 10.07 5.26 9.62
Noc / Est 10.07 5.26 9.62
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 6

Error D1-bg D1-fg D1-all
All / All 11.93 7.35 11.45
All / Est 11.93 7.35 11.45
Noc / All 11.90 7.35 11.41
Noc / Est 11.90 7.35 11.41
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 7

Error D1-bg D1-fg D1-all
All / All 3.75 16.32 6.21
All / Est 3.75 16.32 6.21
Noc / All 3.77 16.32 6.26
Noc / Est 3.77 16.32 6.26
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 8

Error D1-bg D1-fg D1-all
All / All 3.66 4.77 3.87
All / Est 3.66 4.77 3.87
Noc / All 3.66 4.77 3.87
Noc / Est 3.66 4.77 3.87
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 9

Error D1-bg D1-fg D1-all
All / All 4.13 7.43 4.97
All / Est 4.13 7.43 4.97
Noc / All 4.14 7.79 5.05
Noc / Est 4.14 7.79 5.05
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 10

Error D1-bg D1-fg D1-all
All / All 4.24 9.36 5.41
All / Est 4.24 9.36 5.41
Noc / All 4.26 9.36 5.44
Noc / Est 4.26 9.36 5.44
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 11

Error D1-bg D1-fg D1-all
All / All 2.65 5.65 3.19
All / Est 2.65 5.65 3.19
Noc / All 2.64 5.65 3.18
Noc / Est 2.64 5.65 3.18
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 12

Error D1-bg D1-fg D1-all
All / All 1.94 2.63 1.98
All / Est 1.94 2.63 1.98
Noc / All 1.83 2.63 1.89
Noc / Est 1.83 2.63 1.89
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 13

Error D1-bg D1-fg D1-all
All / All 1.72 3.75 1.97
All / Est 1.72 3.75 1.97
Noc / All 1.62 3.75 1.88
Noc / Est 1.62 3.75 1.88
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 14

Error D1-bg D1-fg D1-all
All / All 1.59 5.00 1.65
All / Est 1.59 5.00 1.65
Noc / All 1.47 5.00 1.54
Noc / Est 1.47 5.00 1.54
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 15

Error D1-bg D1-fg D1-all
All / All 4.56 4.46 4.56
All / Est 4.56 4.46 4.56
Noc / All 4.63 4.46 4.62
Noc / Est 4.63 4.46 4.62
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 16

Error D1-bg D1-fg D1-all
All / All 6.61 6.03 6.52
All / Est 6.61 6.03 6.52
Noc / All 6.22 6.03 6.20
Noc / Est 6.22 6.03 6.20
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 17

Error D1-bg D1-fg D1-all
All / All 2.50 1.75 2.43
All / Est 2.50 1.75 2.43
Noc / All 2.22 1.75 2.17
Noc / Est 2.22 1.75 2.17
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 18

Error D1-bg D1-fg D1-all
All / All 8.77 13.79 11.16
All / Est 8.77 13.79 11.16
Noc / All 8.57 13.79 11.07
Noc / Est 8.57 13.79 11.07
This table as LaTeX

Input Image

D1 Result

D1 Error


Test Image 19

Error D1-bg D1-fg D1-all
All / All 2.02 4.34 2.28
All / Est 2.02 4.34 2.28
Noc / All 1.98 4.34 2.25
Noc / Est 1.98 4.34 2.25
This table as LaTeX

Input Image

D1 Result

D1 Error




eXTReMe Tracker