Project Login
Registration No:
Password:
MAIL ALERTS SMS ALERTS
 
     
   
     

Contrast-Independent Curvilinear Structure Detection in Biomedical Images

Platform : DOT NET

IEEE Projects Years : 2012 - 13

Contrast-Independent Curvilinear Structure

Detection in Biomedical Images

ABSTRACT:

Many biomedical applications require detection of curvilinear structures in images and would benefit from automatic or semiautomatic segmentation to allow high-throughput measurements. Here, we propose a contrast-independent approach to identify curvilinear structures based on oriented phase congruency, i.e., the phase congruency tensor (PCT). We show that the proposed method is largely insensitive to intensity variations along

the curve and provides successful detection within noisy regions. The performance of the PCT is evaluated by comparing it with state-of-the-art intensity-based approaches on both synthetic and real biological images.          

EXISTING SYSTEM:

  • A common approach to represent local image structure through a tensor is to consider the first terms of the Taylor series expansion. In the Taylor series, the second-order derivatives are incorporated in the Hessian matrix.
  • In Different practical ways to calculate phase congruency have been proposed .A robust approach was introduced, combining information over several scales and orientations and including a noise compensation term.
  • An interesting approach relies on the use of minimum cost paths, an example of which is the live-wire algorithm. In the live-wire method, the minimum cost path is calculated on a graph representation of the image and using Dijkstra shortest-path algorithm. The shortest path is defined as a path from one node/pixel to another such that the sum of the costs of the arcs on the path is minimized. The use of minimum cost path algorithms and window-constrained global search was suggested. An approach that combines directional matched filtering with an algorithm for finding minimum cost paths was proposed.

 

 

 

 

PROPOSED SYSTEM:

We propose a contrast-independent approach to identify curvilinear structures based on oriented phase congruency, i.e., the phase congruency tensor (PCT). We show that the

Proposed method is largely insensitive to intensity variations along the curve and provides successful detection within noisy regions. The performance of the PCT is evaluated by comparing it with state-of-the-art intensity-based approaches on both synthetic and real biological images.

We have proposed an oriented Gaussian-shaped curve matching filter. The matched filter parameters can be estimated empirically or via an optimization process. Based on this filter, an amplitude-modified second derivative of the Gaussian filter was introduced.

ADVANTAGES OF PROPOSED SYSTEM:

In order to take full advantage of both intensity-and orientation-based cost map components, we also introduce the component based on the vesselness measurement.

HARDWARE AND SOFTWARE REQUIREMENTS         :

HARDWARE REQUIREMENTS:

Processor                     :                       Pentium 4 Cpu 2.40ghz

Ram                             :                       512 Mb Ram

Hard Disk                   :                       40 Gb

Keyboard                    :                       Standard

Monitor                       :                       15”

SOFTWARE REQUIREMENTS:

Front End                    :           C#.Net

Operating System       :           Windows Xp

MODULES DESCRIPTION:

PCT:-

                       The detection of curvilinear structures is particularly affected by variations of intensity contrast within the image. Intensity differences between curvilinear structures and with the back-ground, common to many biomedical imaging applications, cause traditional intensity-based methods to produce widely varying outputs, which, in turn, make it difficult for post-processing methods to delineate the structures. Additionally, the boundaries of low-contrast structures may not be detected by methods based on the image gradient. Therefore, it is essential that the approach is invariant to changes in image intensity and contrast. Changes in intensity may occur because of varying illumination, signal variations, bias fields in magnetic resonance imaging, etc.

we propose a brightness- and contrast-invariant method for curvilinear structure extraction based on the concept of local phase and particularly on a model of phase congruency, which assumes that image features are observed at points in an image where the Fourier components are maximally in phase. Phase-based ridge detectors, such as, have been shown to be able to detect such structures in a largely contrast-independent way. The difference between the phase-based ridge detectors and our Approach is given by the relation between outputs at different orientations. We propose to combine all values of orientation-specific phase congruency in a single tensor, the Eigen-analysis of which can be used as a contrast-independent description of The local structure.

 

PCT VESSELNESS AND NEURITENESS:-

                  As described in section i-d1 and i-d2, piecewise curvilinear segments can be detected by analyzing the relations between eigen values and eigenvectors of the locally calculated hessian. in a similar way, the dominant orientation of the surface repre-senting a curvilinear structure is given by the dominant eigen-vectors. i.e., the eigenvector corresponding to the Eigen-value of largest magnitude. Pct-based vesselness and neurite-ness are calculated . Where the Eigen-values of substitute those of the hessian.

PCT LIVE-WIRE TRACING:-

                       In order to apply the live-wire tracing method discussed in Section I-F, the eight-connected graph representation of the image is constructed. Then, the cost map combining the image intensity cost and the vector-field cost is calculated as defined. We study the use of different cost functions based on previously available intensity-based techniques and our proposed PCT. For the vesselness-based tracing, the component of the

cost map corresponds to the vesselness measurement as de-fined. In order to take full advantage of both intensity-and orientation-based cost map components, we also introduce the component based on the vesselness measurement. we calculated the outputs of relevant previously available methods including a Gaussian matched filter  vesselness and neuriteness. The results of these and the PCT vesselness and neuriteness are presented. The parameters used in these tests were manually optimized to provide the best visual detection, with the PCT-based parameters kept exactly the same as those for their respective intensity-based measures.



NOW GET PROJECTS ! GET TRAINED ! GET PLACED !

IEEE, NON-IEEE, REAL TIME LIVE ACADEMIC PROJECTS,

PROJECTS WITH COMPLETE COURSES,SOFT SKILLS & PLACEMENTS

ALLOVER INDIA & WORLD WIDE,

HOSTEL FACILITY AVAILABLE FOR GIRLS & BOYS SEPARATELY,

CALL: 08985129129 ,  E-Mail Id: support@ascentit.in

REGISTER FOR PROJECTS NOW ! GET DISCOUNT
   
1