Object Removal - Revision history

imported>Projects221: /* Seam Carving for Object Removal */

2013-03-21T14:50:44Z

Seam Carving for Object Removal

← Older revision		Revision as of 14:50, 21 March 2013
Line 230:		Line 230:
	! <Center>[[File:boats-seams.jpg]]</Center>		! <Center>[[File:boats-seams.jpg]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 35: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 36: Target Image</Center>
	\| <Center>Figure 9: Seam Carved Image</Center>		\| <Center>Figure 37: Seam Carved Image</Center>
	\| <Center>Figure 10: Energy Graph</Center>		\| <Center>Figure 38: Energy Graph</Center>
	\| <Center>Figure 10: Selected Seams</Center>		\| <Center>Figure 39: Selected Seams</Center>
	\|}		\|}

Line 245:		Line 245:
	! <Center>[[File:Dogs-seams.jpg]]</Center>		! <Center>[[File:Dogs-seams.jpg]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 40: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 41: Target Image</Center>
	\| <Center>Figure 9: Seam Carved Image</Center>		\| <Center>Figure 42: Seam Carved Image</Center>
	\| <Center>Figure 10: Energy Graph</Center>		\| <Center>Figure 43: Energy Graph</Center>
	\| <Center>Figure 10: Selected Seams</Center>		\| <Center>Figure 44: Selected Seams</Center>
	\|}		\|}

Line 260:		Line 260:
	! <Center>[[File:Bronx-zoo-seams.jpg]]</Center>		! <Center>[[File:Bronx-zoo-seams.jpg]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 45: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 46: Target Image</Center>
	\| <Center>Figure 9: Seam Carved Image</Center>		\| <Center>Figure 47: Seam Carved Image</Center>
	\| <Center>Figure 10: Energy Graph</Center>		\| <Center>Figure 48: Energy Graph</Center>
	\| <Center>Figure 10: Selected Seams</Center>		\| <Center>Figure 49: Selected Seams</Center>
	\|}		\|}

Line 277:		Line 277:
	! <Center>[[File:Movie-show-seams.jpg]]</Center>		! <Center>[[File:Movie-show-seams.jpg]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 50: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 51: Target Image</Center>
	\| <Center>Figure 9: Seam Carved Image</Center>		\| <Center>Figure 52: Seam Carved Image</Center>
	\| <Center>Figure 10: Energy Graph</Center>		\| <Center>Figure 53: Energy Graph</Center>
	\| <Center>Figure 10: Selected Seams</Center>		\| <Center>Figure 54: Selected Seams</Center>
	\|}		\|}

Line 292:		Line 292:
	! <Center>[[File:Crowd-seams.jpg]]</Center>		! <Center>[[File:Crowd-seams.jpg]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 55: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 56: Target Image</Center>
	\| <Center>Figure 9: Seam Carved Image</Center>		\| <Center>Figure 57: Seam Carved Image</Center>
	\| <Center>Figure 10: Energy Graph</Center>		\| <Center>Figure 58: Energy Graph</Center>
	\| <Center>Figure 10: Selected Seams</Center>		\| <Center>Figure 59: Selected Seams</Center>
	\|}		\|}

imported>Projects221: /* Worst Cases */

2013-03-21T14:49:24Z

Worst Cases

← Older revision		Revision as of 14:49, 21 March 2013
Line 208:		Line 208:
	! <Center>[[File:st_partialP2.png]]</Center>		! <Center>[[File:st_partialP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 15: Original Image</Center>		\| <Center>Figure 31: Original Image</Center>
	\| <Center>Figure 16: Target Image</Center>		\| <Center>Figure 32: Target Image</Center>
	\| <Center>Figure 17: Inpainted Image</Center>		\| <Center>Figure 33: Inpainted Image</Center>
	\| <Center>Figure 18: Confidence and Data Terms</Center>		\| <Center>Figure 34: Confidence and Data Terms</Center>
	\|}		\|}

imported>Projects221: /* Seam Carving for Object Removal */

2013-03-21T14:49:04Z

Seam Carving for Object Removal

← Older revision		Revision as of 14:49, 21 March 2013
Line 219:		Line 219:
	We analyzed the seam carving object removal algorithm by visually analyzing the results on a series of test images. These images varied in content, color, foreground and background composition, as well as texture, and target object size. We performed basic visual saliency test to see if the altercation were immediately apparent. We also performed more detailed analysis to determine where and why the algorithm left visual artifacts.		We analyzed the seam carving object removal algorithm by visually analyzing the results on a series of test images. These images varied in content, color, foreground and background composition, as well as texture, and target object size. We performed basic visual saliency test to see if the altercation were immediately apparent. We also performed more detailed analysis to determine where and why the algorithm left visual artifacts.

	The algorithm performed well for both small and medium sized target object, but left the image unrecognizable when removed large objects. Large objects~~, like in figure (22),~~ produced poor results largely because the large number of seams that had to be removed in order to remove the object entirely. The algorithm also showed that if could work well in both cluttered and plain backgrounds, but showed easily recognizable errors when the image contain patterned backgrounds. We also produced visual displays of the images energy functions and selected visual seams, in order to better visualize how the content was being manipulated.		The algorithm performed well for both small and medium sized target object, but left the image unrecognizable when removed large objects. Large objects produced poor results largely because the large number of seams that had to be removed in order to remove the object entirely. The algorithm also showed that if could work well in both cluttered and plain backgrounds, but showed easily recognizable errors when the image contain patterned backgrounds. We also produced visual displays of the images energy functions and selected visual seams, in order to better visualize how the content was being manipulated.

	==== '''Best Cases''' ====		==== '''Best Cases''' ====

imported>Projects221: /* Best Cases */

2013-03-21T14:48:32Z

Best Cases

← Older revision		Revision as of 14:48, 21 March 2013
Line 127:		Line 127:
	! <Center>[[File:bungeeP2.png]]</Center>		! <Center>[[File:bungeeP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 7: Original Image</Center>		\| <Center>Figure 8: Original Image</Center>
	\| <Center>Figure 8: Target Image</Center>		\| <Center>Figure 9: Target Image</Center>
	\| <Center>Figure 9: Inpainted Image</Center>		\| <Center>Figure 10: Inpainted Image</Center>
	\| <Center>Figure 10: Confidence and Data Terms</Center>		\| <Center>Figure 11: Confidence and Data Terms</Center>
	\|}		\|}

Line 140:		Line 140:
	! <Center>[[File:stP2.png]]</Center>		! <Center>[[File:stP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 11: Original Image</Center>		\| <Center>Figure 12: Original Image</Center>
	\| <Center>Figure 12: Target Image</Center>		\| <Center>Figure 13: Target Image</Center>
	\| <Center>Figure 13: Inpainted Image</Center>		\| <Center>Figure 14: Inpainted Image</Center>
	\| <Center>Figure 14: Confidence and Data Terms</Center>		\| <Center>Figure 15: Confidence and Data Terms</Center>
	\|}		\|}

Line 153:		Line 153:
	! <Center>[[File:bwP2.png]]</Center>		! <Center>[[File:bwP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 15: Original Image</Center>		\| <Center>Figure 16: Original Image</Center>
	\| <Center>Figure 16: Target Image</Center>		\| <Center>Figure 17: Target Image</Center>
	\| <Center>Figure 17: Inpainted Image</Center>		\| <Center>Figure 18: Inpainted Image</Center>
	\| <Center>Figure 18: Confidence and Data Terms</Center>		\| <Center>Figure 19: Confidence and Data Terms</Center>
	\|}		\|}

Line 166:		Line 166:
	! <Center>[[File:dogsP2.png]]</Center>		! <Center>[[File:dogsP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 15: Original Image</Center>		\| <Center>Figure 20: Original Image</Center>
	\| <Center>Figure 16: Target Image</Center>		\| <Center>Figure 21: Target Image</Center>
	\| <Center>Figure 17: Inpainted Image</Center>		\| <Center>Figure 22: Inpainted Image</Center>
	\| <Center>Figure 18: Confidence and Data Terms</Center>		\| <Center>Figure 23: Confidence and Data Terms</Center>
	\|}		\|}

Line 179:		Line 179:
	! <Center>[[File:movieP2.png]]</Center>		! <Center>[[File:movieP2.png]]</Center>
	\|-		\|-
	\| <Center>Figure 15: Original Image</Center>		\| <Center>Figure 24: Original Image</Center>
	\| <Center>Figure 16: Target Image</Center>		\| <Center>Figure 25: Target Image</Center>
	\| <Center>Figure 17: Inpainted Image</Center>		\| <Center>Figure 26: Inpainted Image</Center>
	\| <Center>Figure 18: Confidence and Data Terms</Center>		\| <Center>Figure 27: Confidence and Data Terms</Center>
	\|}		\|}

Line 191:		Line 191:
	! <Center>[[File:sign3.png]]</Center>		! <Center>[[File:sign3.png]]</Center>
	\|-		\|-
	\| <Center>Figure 15: Original Image</Center>		\| <Center>Figure 28: Original Image</Center>
	\| <Center>Figure 16: Target Image</Center>		\| <Center>Figure 29: Target Image</Center>
	\| <Center>Figure 17: Inpainted Image</Center>		\| <Center>Figure 30: Inpainted Image</Center>
	\|}		\|}
	<Center>''Note: This image is copied from [2], and was not run on our machines''</Center>		<Center>''Note: This image is copied from [2], and was not run on our machines''</Center>

imported>Projects221: /* Exemplar-Based Inpainting */

2013-03-21T14:47:11Z

Exemplar-Based Inpainting

← Older revision		Revision as of 14:47, 21 March 2013
Line 51:		Line 51:

	<Center>[[File:ConfAndData.png]]</Center>		<Center>[[File:ConfAndData.png]]</Center>
	<Center>Figure 7: Effects of data and conﬁdence terms[3]</Center>		<Center>Figure 6: Effects of data and conﬁdence terms[3]</Center>

	2. The second step involves propagating texture and structure information. Once all the priorities have been calculated, we take the patch with the highest priority and fill it with information from the source region. The information from the source region is directly sampled as opposed to being propagated using diffusion. Diffusion leads to image smoothing and as a result a blurry fill-in, especially of large regions.		2. The second step involves propagating texture and structure information. Once all the priorities have been calculated, we take the patch with the highest priority and fill it with information from the source region. The information from the source region is directly sampled as opposed to being propagated using diffusion. Diffusion leads to image smoothing and as a result a blurry fill-in, especially of large regions.
Line 67:		Line 67:
	<Center>[[File:InpaintingAlgorithm.png]]</Center>		<Center>[[File:InpaintingAlgorithm.png]]</Center>

	<Center>Figure 6: Region Filling Algorithm[3]</Center>		<Center>Figure 7: Region Filling Algorithm[3]</Center>

	We analyzed the exemplar-based inpainting algorithm by running it on a set of images while varying the size and location of the target region on a subset of these images. This allowed us to determine the impact of the fill region as well as it allowed us to determine on what set of images this algorithm worked well for and for what set of images it didn’t work well for. We measured the success/failure of the algorithm visually by either assigning it a pass or fail. The images we used were a combination of synthetic and real photographs with a variety of textures as backgrounds.		We analyzed the exemplar-based inpainting algorithm by running it on a set of images while varying the size and location of the target region on a subset of these images. This allowed us to determine the impact of the fill region as well as it allowed us to determine on what set of images this algorithm worked well for and for what set of images it didn’t work well for. We measured the success/failure of the algorithm visually by either assigning it a pass or fail. The images we used were a combination of synthetic and real photographs with a variety of textures as backgrounds.

imported>Projects221: /* Worst Cases */

2013-03-21T14:45:39Z

Worst Cases

← Older revision		Revision as of 14:45, 21 March 2013
Line 289:		Line 289:
	! <Center>[[File:Crowd_temp.png]]</Center>		! <Center>[[File:Crowd_temp.png]]</Center>
	! <Center>[[File:Crowd-removed.jpg]]</Center>		! <Center>[[File:Crowd-removed.jpg]]</Center>
	! <Center>[[File:Crowd-energy.~~png~~]]</Center>		! <Center>[[File:Crowd-energy.jpg]]</Center>
	! <Center>[[File:Crowd-seams.jpg]]</Center>		! <Center>[[File:Crowd-seams.jpg]]</Center>
	\|-		\|-

imported>Projects221: /* Worst Cases */

2013-03-21T14:44:24Z

Worst Cases

@@ Line 276: / Line 276: @@
 ! <Center>[[File:Energy-movie.png]]</Center>
 ! <Center>[[File:Movie-show-seams.jpg]]</Center>
 |-
 | <Center>Figure 7: Original Image</Center>

imported>Projects221: /* Results/Analysis */

2013-03-21T14:34:07Z

Results/Analysis

← Older revision		Revision as of 14:34, 21 March 2013
Line 221:		Line 221:
	The algorithm performed well for both small and medium sized target object, but left the image unrecognizable when removed large objects. Large objects, like in figure (22), produced poor results largely because the large number of seams that had to be removed in order to remove the object entirely. The algorithm also showed that if could work well in both cluttered and plain backgrounds, but showed easily recognizable errors when the image contain patterned backgrounds. We also produced visual displays of the images energy functions and selected visual seams, in order to better visualize how the content was being manipulated.		The algorithm performed well for both small and medium sized target object, but left the image unrecognizable when removed large objects. Large objects, like in figure (22), produced poor results largely because the large number of seams that had to be removed in order to remove the object entirely. The algorithm also showed that if could work well in both cluttered and plain backgrounds, but showed easily recognizable errors when the image contain patterned backgrounds. We also produced visual displays of the images energy functions and selected visual seams, in order to better visualize how the content was being manipulated.

			==== '''Best Cases''' ====
	{\| border="1" cellspacing="0" cellpadding="5" align="center"		{\| border="1" cellspacing="0" cellpadding="5" align="center"
	\|-		\|-
Line 266:		Line 267:
	\|}		\|}

			==== '''Worst Cases''' ====

	{\| border="1" cellspacing="0" cellpadding="5" align="center"		{\| border="1" cellspacing="0" cellpadding="5" align="center"

imported>Projects221: /* Seam Carving for Object Removal */

2013-03-21T14:29:34Z

Seam Carving for Object Removal

@@ Line 258: / Line 258: @@
 ! <Center>[[File:Bronx-zoo-energy.jpg]]</Center>
 ! <Center>[[File:Bronx-zoo-seams.jpg]]</Center>
 |-
 | <Center>Figure 7: Original Image</Center>

imported>Projects221: /* Seam Carving for Object Removal */

2013-03-21T14:25:04Z

Seam Carving for Object Removal

@@ Line 223: / Line 223: @@
 {| border="1" cellspacing="0" cellpadding="5" align="center"
 |-
-! <Center>[[File:b0.png]]</Center>
+! <Center>[[File:Object-removal-before.jpg]]</Center>
-! <Center>[[File:b1.png]]</Center>
+! <Center>[[File:Object-removal-after.png]]</Center>
-! <Center>[[File:bA.png]]</Center>
+! <Center>[[File:boat-removal.jpg]]</Center>
-! <Center>[[File:bungeeP2.png]]</Center>
+! <Center>[[File:boats-energy.jpg]]</Center>
 |-
 | <Center>Figure 7: Original Image</Center>