SoC2007 project Feature Matching - Revision history

Erik Krause at 16:48, 20 November 2007

2007-11-20T16:48:53Z

Bruno: categorised

2007-04-03T21:32:33Z

categorised

Mann: /* Students planning to apply */

2007-03-26T05:42:44Z

‎Students planning to apply

Paler: /* Project parts short description */

2007-03-25T14:53:58Z

‎Project parts short description

Paler: /* Project parts short description */

2007-03-25T14:47:07Z

‎Project parts short description

Paler: /* Project parts short description */

2007-03-25T14:40:48Z

‎Project parts short description

Paler: /* Detailed timeschedule */

2007-03-25T14:29:25Z

‎Detailed timeschedule

Paler: /* Details */

2007-03-25T11:02:05Z

‎Details

Paler: /* Details */

2007-03-25T11:01:39Z

‎Details

Paler: /* Detailed timeschedule */

2007-03-25T10:54:22Z

‎Detailed timeschedule

@@ Line 26: / Line 26: @@
 Another part of the project is  RANSAC (Random Sample Consensus Algorithm). This is an algorithm used for robust fitting the models in the presence of many data outliers.
-The wikipedia[http://en.wikipedia.org/wiki/RANSAC] article describing this algorithm is offering its pseudo code. There is also a paper[http://www.tu-chemnitz.de/etit/proaut/paperdb/download/fischler81.pdf] describing this algorithm which offers a helpful and proper understanding and implementation.
+The [[wikipedia:RANSAC]] article describing this algorithm is offering its pseudo code. There is also a paper[http://www.tu-chemnitz.de/etit/proaut/paperdb/download/fischler81.pdf] describing this algorithm which offers a helpful and proper understanding and implementation.
 Cover trees together with RANSAC can return good results, these results however can be improved by using heuristics. Image stitching can be improved by using information that is associated with the images. This information can be found in the EXIF structure that many cameras can append to the shots taken. There are some habits involved with taking panorama images. Panoramas can be built out of multiple images that are organized in one or multiple rows. Still some panoramas are difficult to automatically stitch because of the similarity between the feature points. That is why, knowing or assuming the order the images were taken can improve the results of this final and important stitching step. Determining the habits of photographers and introducing it in an algorithm is an important part of this project.

← Older revision		Revision as of 21:32, 3 April 2007
Line 144:		Line 144:
	==Resources==		==Resources==
	* [http://hunch.net/~jl/projects/cover_tree/cover_tree.html Fast nearest neighbour matching using cover trees]		* [http://hunch.net/~jl/projects/cover_tree/cover_tree.html Fast nearest neighbour matching using cover trees]
		+
		+	[[Category:Community:Project]]

@@ Line 140: / Line 140: @@
 ==Students planning to apply==
 * [[User:Paler | Alexandru Paler]]
 ==Resources==
 * [http://hunch.net/~jl/projects/cover_tree/cover_tree.html Fast nearest neighbour matching using cover trees]

@@ Line 26: / Line 26: @@
 Another part of the project is  RANSAC (Random Sample Consensus Algorithm). This is an algorithm used for robust fitting the models in the presence of many data outliers.
-The wikipedia[http://en.wikipedia.org/wiki/RANSAC] article describing this algorithm is offering its pseudo code. There is also a paper[http://www.tu-chemnitz.de/etit/proaut/paperdb/download/fischler81.pdf] describing this algorithm which offers a helpful and proper understanding and implementation
+The wikipedia[http://en.wikipedia.org/wiki/RANSAC] article describing this algorithm is offering its pseudo code. There is also a paper[http://www.tu-chemnitz.de/etit/proaut/paperdb/download/fischler81.pdf] describing this algorithm which offers a helpful and proper understanding and implementation.
 Cover trees together with RANSAC can return good results, these results however can be improved by using heuristics. Image stitching can be improved by using information that is associated with the images. This information can be found in the EXIF structure that many cameras can append to the shots taken. There are some habits involved with taking panorama images. Panoramas can be built out of multiple images that are organized in one or multiple rows. Still some panoramas are difficult to automatically stitch because of the similarity between the feature points. That is why, knowing or assuming the order the images were taken can improve the results of this final and important stitching step. Determining the habits of photographers and introducing it in an algorithm is an important part of this project.

@@ Line 30: / Line 30: @@
 Cover trees together with RANSAC can return good results, these results however can be improved by using heuristics. Image stitching can be improved by using information that is associated with the images. This information can be found in the EXIF structure that many cameras can append to the shots taken. There are some habits involved with taking panorama images. Panoramas can be built out of multiple images that are organized in one or multiple rows. Still some panoramas are difficult to automatically stitch because of the similarity between the feature points. That is why, knowing or assuming the order the images were taken can improve the results of this final and important stitching step. Determining the habits of photographers and introducing it in an algorithm is an important part of this project.
-The nearest neighbour matching and RANSAC outlier pruning could be combined into an integrated algorithm that considers both the feature vector as well as the feature position at the same time. This requires knowledge of the focal length and distortion parameters of the images, which could be read from a lens database file.
+The nearest neighbour matching and RANSAC outlier pruning could be combined into an integrated algorithm that considers both the feature vector as well as the feature position at the same time. This requires knowledge of the focal length and distortion parameters of the images, which could be read from a lens database file. This combined algorithm could reduce considerably the time required for the feature matching.
 ===EXIF heuristics===

@@ Line 23: / Line 23: @@
 ===Project parts short description===
-The first part of the matching will use a nearest matching algorithm, more precisely cover trees. A cover tree is a data structure helpful in calculating the nearest neighbor of points given only a measure. There are papers documenting this data structure and the code implementation [http://hunch.net/~jl/projects/cover_tree/cover_tree.html]. To implement this algorithm represents an advantage.It is saving time for implementation and testing.
+The first part of the matching will use a nearest neighbor matching algorithm, more precisely cover trees. A cover tree is a data structure helpful in calculating the nearest neighbor of points given only a measure. There are papers documenting this data structure and the code implementation [http://hunch.net/~jl/projects/cover_tree/cover_tree.html]. To implement this algorithm represents an advantage.It is saving time for implementation and testing.
 Another part of the project is  RANSAC (Random Sample Consensus Algorithm). This is an algorithm used for robust fitting the models in the presence of many data outliers.
@@ Line 29: / Line 29: @@
 Cover trees together with RANSAC can return good results, these results however can be improved by using heuristics. Image stitching can be improved by using information that is associated with the images. This information can be found in the EXIF structure that many cameras can append to the shots taken. There are some habits involved with taking panorama images. Panoramas can be built out of multiple images that are organized in one or multiple rows. Still some panoramas are difficult to automatically stitch because of the similarity between the feature points. That is why, knowing or assuming the order the images were taken can improve the results of this final and important stitching step. Determining the habits of photographers and introducing it in an algorithm is an important part of this project.
 ===EXIF heuristics===

@@ Line 76: / Line 76: @@
 Planned actions:
-* Getting used to panorama making and finding the associated problems in a practical way
+* Getting used to panorama making and finding the associated problems in a practical way;
 * Reading scientific papers describing the problems encountered by other people and the way they solved them;
 * Discussions with the mentor about the already existing code and ways to use or improve it;

@@ Line 15: / Line 15: @@
 Standard algorithms and data structures combined with heuristic methods will form the core of the suggested project solution.
-The Recognizing panoramas paperwork written by M. Brown and D.G. Lowe [http://www.cs.ubc.ca/~lowe/papers/brown03.pdf] describes the process of panorama creation from the algorithms' perspective.
+The ''Recognizing panoramas'' paperwork written by M. Brown and D.G. Lowe [http://www.cs.ubc.ca/~lowe/papers/brown03.pdf] describes the process of panorama creation from the algorithms' perspective.
 The first step implies that the RANSAC algorithm is used to compute the matches between the images. The following step is to compute if the matches are really correct by using a probabilistic model, also described in the above mentioned document. The third step is to join the images one by one. The result image that is going to be joined is the one with the most matches available.

@@ Line 114: / Line 114: @@
 *Improving the code and the results.
-''All the dates are less exact, but the time frames allowed should indicate the way the project is about to be completed.''
+In the open-source projects seems to be a never-ending process of software development, in which new features are added to the software and corrections are made. This process is not to be characterized as maintenance though and is analogous to what may be referred to as versioning (evolution). That is why, tests will be done also during the implementation part. To be able to get a correct evaluation and to have valid tests, a test set is going to be prepared at the beginning of the research part. This test set is going to contain images relevant for the project goal.
 ==Deliverables==