Depth Mapping Algorithm Performance Analysis - Revision history

imported>Student2017: /* Test 1: */

2017-12-16T14:34:39Z

Test 1:

← Older revision		Revision as of 14:34, 16 December 2017
Line 115:		Line 115:

	=====Test 1: =====		=====Test 1: =====
	One image was focused on for our analysis of these algorithms. The same block size ~~would~~ be used for all test in order to compare the algorithms alone. Ideally all resulting disparity maps will be compared with the "ground truth" numerically and ~~quantify~~ a performance based on error, but for now resulting disparity maps ~~will~~ be visually compared.		One image was focused on for our analysis of these algorithms. The same block size was be used for all test in order to compare the algorithms alone. Ideally all resulting disparity maps will be compared with the "ground truth" numerically and a performance based on error could be quantified, but for now resulting disparity maps would be visually compared.

	=====Test 2: =====		=====Test 2: =====

imported>Student2017: /* Block/Window Matching */

2017-12-16T14:31:35Z

Block/Window Matching

← Older revision		Revision as of 14:31, 16 December 2017
Line 72:		Line 72:

	===Block/Window Matching===		===Block/Window Matching===
	~~To analyse~~ an image, it ~~is broken~~ down into subsections called blocks or windows, and ~~comparisons are performed on~~ these subsections.		Analysis of an image is performed by breaking it down into subsections called blocks or windows and these subsections are used for comparison.
	[[File:og_pic.png\|thumb\|center\|300px\|Figure 8: Example original picture, left stereo image [5]]]		[[File:og_pic.png\|thumb\|center\|300px\|Figure 8: Example original picture, left stereo image [5]]]

imported>Student2017: /* Middlebury Dataset */

2017-12-16T14:27:32Z

Middlebury Dataset

← Older revision		Revision as of 14:27, 16 December 2017
Line 65:		Line 65:
	[[File:block_d.png\|thumb\|center\|500px\|Figure 7: Block diagram of how highly accurate "ground truth" was made [4]]]		[[File:block_d.png\|thumb\|center\|500px\|Figure 7: Block diagram of how highly accurate "ground truth" was made [4]]]

	~~All~~ this information ~~and helped bypass the~~ issues with the SYNS dataset. ~~The~~ "perfectly" rectified stereo images from the dataset were used ~~easy the process~~ and ~~see~~ direct comparisons between ~~methods~~ mentioned in the later sections.		With all this information, issues seen with the SYNS dataset could be easily bypassed. Therefore the "perfectly" rectified stereo images from the dataset were used, providing ease of implementation and direct comparisons between algorithms mentioned in the later sections.

	== Methods: Block Matching and Similarity Metrics ==		== Methods: Block Matching and Similarity Metrics ==

imported>Student2017: /* Block/Window Matching */

2017-12-16T12:00:29Z

Block/Window Matching

← Older revision		Revision as of 12:00, 16 December 2017
Line 78:		Line 78:
	{\|style="margin: 0 auto;"		{\|style="margin: 0 auto;"
	\|[[File:left_crop.png\|thumb\|center\|400px\|Figure 9a: Zoomed-in region with block selected for left stereo image. Left image is used as reference]]		\|[[File:left_crop.png\|thumb\|center\|400px\|Figure 9a: Zoomed-in region with block selected for left stereo image. Left image is used as reference]]
	\|[[File:right_crop.png\|thumb\|center\|400px\|Figure 9b: Same zoomed-in region in right stereo image with horizontal sweep bounds shown and original block position]]		\|[[File:right_crop.png\|thumb\|center\|400px\|Figure 9b: Same zoomed-in region in right stereo image with horizontal sweep bounds shown and original block position in white]]
	\|}		\|}

	The		The reference block from the left stereo image now has to be compared with the acquired block from the right stereo image. To do this the color image is translated to grayscale and each pixel the within the block now can have a numerical intensity value, Figure 10. Depending on algorithm these values are used to evaluate a correlation between the blocks and this can lead to calculating the disparity between the two images at that point in the reference image.
	[[File:d_cal.png\|thumb\|center\|400px\|Figure 8: Example original picture, left stereo image [5]]]		[[File:d_cal.png\|thumb\|center\|400px\|Figure 10: Example original picture, left stereo image [5]]]

	===SAD===		===SAD===

imported>Student2017: /* Test 2: */

2017-12-16T11:47:26Z

Test 2:

← Older revision		Revision as of 11:47, 16 December 2017
Line 118:		Line 118:

	=====Test 2: =====		=====Test 2: =====
	Since the block size used would have an effect accuracy of the resulting disparity map, the SAD algorithm was used to test the effect or block size variation. Apart from just block size, Semi-Global Matching (SGM) was also investigate so see its effect over different sizing. SGM is based on the idea pixelwise matching of Mutual Information and approximating a global, 2D smoothness constraint by combining many 1D constraints [6].		Since the block size used would have an effect accuracy of the resulting disparity map, the SAD algorithm was used to test the effect or block size variation. Apart from just block size, Semi-Global Matching (SGM) was also investigate so see its effect over different sizing. SGM is based on the idea pixelwise matching of Mutual Information and approximating a global, 2D smoothness constraint by combining many 1D constraints [10].

	== Results ==		== Results ==

imported>Student2017: /* References */

2017-12-16T11:47:14Z

References

← Older revision		Revision as of 11:47, 16 December 2017
Line 182:		Line 182:

	[9] Adams, W.J., Elder, J.H., Graf, E.W., Leyland, J., Lugtigheid, A.J., Muryy, A. (2016). The Southampton-York Natural Scenes (SYNS) dataset: Statistics of surface attitude. Scientific Reports, 6, 35805.		[9] Adams, W.J., Elder, J.H., Graf, E.W., Leyland, J., Lugtigheid, A.J., Muryy, A. (2016). The Southampton-York Natural Scenes (SYNS) dataset: Statistics of surface attitude. Scientific Reports, 6, 35805.

			[10] Heiko Hirschmuller, "Accurate and Efficient Stereo Processing by Semi-Global Matching and Mutual Information," IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2005.

	== Appendix I ==		== Appendix I ==

imported>Student2017: /* Test 2: */

2017-12-16T11:45:04Z

Test 2:

← Older revision		Revision as of 11:45, 16 December 2017
Line 118:		Line 118:

	=====Test 2: =====		=====Test 2: =====
	Since the block size used would have an effect accuracy of the resulting disparity map, the SAD algorithm was used to test the effect or block size variation. Apart from just block size,		Since the block size used would have an effect accuracy of the resulting disparity map, the SAD algorithm was used to test the effect or block size variation. Apart from just block size, Semi-Global Matching (SGM) was also investigate so see its effect over different sizing. SGM is based on the idea pixelwise matching of Mutual Information and approximating a global, 2D smoothness constraint by combining many 1D constraints [6].

	== Results ==		== Results ==

imported>Student2017: /* Disparity and Depth */

2017-12-16T11:42:01Z

Disparity and Depth

← Older revision		Revision as of 11:42, 16 December 2017
Line 14:		Line 14:


	Disparity and depth can be related by the following equation (where x-x' is disparity, z is depth, f is the focal length, and B is the interocular distance). Since depth reduced to a scaled quantity of disparity, the accuracy of the disparity could be said to be the more important calculation and disparity calculation could be interchangeable with depth calculation ~~<ref>http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_calib3d/py_depthmap/py_depthmap.html</ref>~~.		Disparity and depth can be related by the following equation (where x-x' is disparity, z is depth, f is the focal length, and B is the interocular distance). Since depth reduced to a scaled quantity of disparity, the accuracy of the disparity could be said to be the more important calculation and disparity calculation could be interchangeable with depth calculation.
	:::::::::::::::::::::::<math>x - x' = \frac{Bf}{z}</math>		:::::::::::::::::::::::<math>x - x' = \frac{Bf}{z}</math>

imported>Student2017: /* SYNS Dataset */

2017-12-16T11:41:23Z

SYNS Dataset

← Older revision		Revision as of 11:41, 16 December 2017
Line 35:		Line 35:
	===SYNS Dataset===		===SYNS Dataset===

	The Southampton-York Natural Scenes (SYNS) dataset contains image and 3D range data measured from different rural and urban locations. Each sample contains the following data:		The Southampton-York Natural Scenes (SYNS) dataset contains image and 3D range data measured from different rural and urban locations [9]. Each sample contains the following data:
	*LiDAR depth information (360 x 135 degree field of view)		*LiDAR depth information (360 x 135 degree field of view)
	*Panoramic HDR image captured with a SpheroCam (Nikkor fish-eye lens) (360 x 180 degree image)		*Panoramic HDR image captured with a SpheroCam (Nikkor fish-eye lens) (360 x 180 degree image)

imported>Student2017: /* Image Rectification */

2017-12-16T11:40:58Z

Image Rectification

← Older revision		Revision as of 11:40, 16 December 2017
Line 27:		Line 27:
	* First we detect SURF keypoints (a scale and rotation invariant feature detector) in each stereo image and extract the feature vectors of each keypoint.		* First we detect SURF keypoints (a scale and rotation invariant feature detector) in each stereo image and extract the feature vectors of each keypoint.
	* We then find matching keypoints between the images using a similarity metric. The uncalibrated rectification used by MATLAB uses the sum of absolute differences metric.		* We then find matching keypoints between the images using a similarity metric. The uncalibrated rectification used by MATLAB uses the sum of absolute differences metric.
	* We then remove outliers (incorrect matches) using an epipolar constraint. Looking at Figure 3, this means that for a keypoint x on the left image, the matching keypoint on the right image must lie on the corresponding epipolar line defined by the intersection of the epipolar plane and the image plane.		* We then remove outliers (incorrect matches) using an epipolar constraint. Looking at Figure 3, this means that for a keypoint x on the left image, the matching keypoint on the right image must lie on the corresponding epipolar line defined by the intersection of the epipolar plane and the image plane. [8]
	* Now using the vector distance between the remaining correct matches, we must compute a 2D projective geometric transformation to apply to one or both of the stereo images. After transformation, the stereo images should no longer have any vertical displacement.		* Now using the vector distance between the remaining correct matches, we must compute a 2D projective geometric transformation to apply to one or both of the stereo images. After transformation, the stereo images should no longer have any vertical displacement.