-
1.
TOMSY
-
(Portfolio)
-
... Temporal Constraints in Variational Correspondence Methods
Segmentation and Object Recognition Results
Following figure shows results for the segmentation algorithm that I deviced while working on ...
-
Created on 07 February 2016
-
2.
Evaporator Automation
-
(2014)
-
Back to the roots! Before obtaining PhD degree in the area of machine vision I designed automation systems. This year I "returned back to the roots" (for a shortwhile only) and designed a complete automation ...
-
Created on 23 November 2014
-
3.
Non-linear Image Diffusion and Efficient Solver
-
(Tutorials)
-
... values are roughly the same. Therefore, we blend image information inside objects while borders (also known as edges) are left intact. As an example, suppose we have two different objects in the image, ...
-
Created on 27 April 2014
-
4.
Collaborators
-
(Collaboration)
-
... Middle East Technical University Inonu Bulvari, 06531Ankara (TURKEY)
Sinan Kalkan's homepage
We met with Sinan while working in the DRIVSCO project. We wrote a scientific acticle with Sinan and ...
-
Created on 20 November 2012
-
5.
A Method for Sparse Disparity Densification using Voting Mask Propagation
-
(Publications)
-
... intensity means more votes, while low intensity means less votes. Black represents 0 votes.
By placing a mask (that best syntonizes with the image gradient direction) on top of each disparity estimation, ...
-
Created on 16 November 2012
-
6.
Spatial and Temporal Constraints in Variational Correspondence Methods
-
(Publications)
-
... WO apriori shows results without using a spatial constraint while Spatial displays results using a spatial constraint. As it can be observed from the video, results with a spatial constraint are far better ...
-
Created on 16 November 2012
-
7.
Disparity Disambiguation by Fusion of Signal- and Symbolic-level Information
-
(Publications)
-
... from the cameras while WO guide refers to results without fusion and W guide refers to results obtained using the method described in the paper. As it can be observed from the video, stereo disparity results ...
-
Created on 16 November 2012
-
8.
Geometric Active Contour Code
-
(Code samples)
-
... \( I \) is the image and \( c \) is a `balloon' force constant.
Segmentation results for an image from the DRIVSO project.
Model 1. is the geometric model, while model 2. is the geodesic model. ...
-
Created on 11 November 2012
-
9.
Level-Set Based Disparity Map Segmentation Code
-
(Code samples)
-
... don't want the interface to be "arbitrarily" long. The theory behind the model is covered briefly here, while both the mathematical model and the exact algorithm implemented in the Matlab code is covered ...
-
Created on 09 May 2012
-
10.
Optical Flow Code
-
(Code samples)
-
... is the coarse-to-fine loop
WHILE( \(s=scl:-1:1\) )
\(I_0 = Iscl_0\, \, I_1 = Iscl_1\ \)
Approximate derivatives for \(I_0\) and \(I_1\): \(\dfrac\), \(\dfrac}\), \(\dfrac} \)
//Solve for new ...
-
Created on 09 May 2012
-
11.
Stereo Disparity Code
-
(Code samples)
-
... loop
WHILE( \(s=scl:-1:1 \) )
\(I_0 =Iscl_0\ \), \(I_1 =Iscl_1\; \)
//Warping loop
WHILE( \( fstLoop \))
//Warp image as per \(u \) and \(v \)
\(I_^w = warp( I_, u, v ); \)
Approximate ...
-
Created on 09 May 2012
-
12.
More information
-
(Publications)
-
... error. The figure below shows a set of 8 voting masks. Intensity represents the number of votes each position gets. High intensity means more votes, while low intensity means less votes. Black represents ...
-
Created on 09 May 2012
-
13.
Segmentation
-
(Topics)
-
... (homogeneous) as per the used similarity metric (e.g. similar color in the case of RGB/HSV images), while the boundary separating the segments is called the interface. Therefore, using explicit curve representation, ...
-
Created on 09 May 2012
-
14.
Optical Flow
-
(Topics)
-
... Gibson in 1940s. In computer vision optical flow typically refers to movement observed in the image plane defined by a vector \( [u,v] \) where the first component describes horizontal movement while the ...
-
Created on 09 May 2012
-
15.
Stereo Disparity
-
(Topics)
-
... are the 'left-' and 'right' camera centres, \(X\) is the point of interest in 3D world, while \(x\) and \(x^\prime\) are the images of \(X\) as seen in the respective cameras. As it can be understood, this ...
-
Created on 09 May 2012
-
16.
Research
-
(Research)
-
... topics include stereo disparity, optical flow and segmentation. Disparity and optical flow are low-level techniques while segmentation is considered a middle-level technique.
3D-reconstruction and ...
-
Created on 09 May 2012
-
17.
ATC
-
(Affiliations)
-
...
Main Research Interests
General research interests of Prof. Ros's group are related to real-time image processing, HDL and hardware design, reconfigurable systems and machine learning, while I am more ...
-
Created on 09 May 2012
