MOSAICING IMAGES_sjec

MOSAICING IMAGES USING DIRECT METHOD

Kavya TantryKirthi Bhat K

SJEC,Vamanjoor

CONTENTS:1. INTRODUCTION2. IMAGE REGISTRATION3. IMAGE WARPING4. IMAGE COMPOSITING5. APPLICATIONS6. CONCLUSION7. REFRENCES

1.INTRODUCTION: Mosaicing is a process of assembling images

taken at different view/time to obtain high resoluted image.

Mosaicing is done in two methods:1. Direct method2. Feature based method

Different process involved in obtaining a mosaiced image are:

1. Image registration2. Image wrapping3. Image compositing

2.IMAGE REGISTRATION: It a process where images taken from different

viewpoints or at different time are registered in databases of the system.

It geometrically aligns the source and the target image.

Its application is divided into four groups:1. Multiview analysis2. Multitemporal analysis3. Multimodal analysis4. Scene to model registration

Steps involved in image registration:

1. Feature detection

2. Feature matching

3. Transform model estimation

4. Image resampling and transformation

3.IMAGE WARPING:

It is a process of overlapping of images taken at different time or from different viewpoints.

Working is done using geometric transformation having mathematic background.

4.IMAGE COMPOSITING:

It mainly involves image blending process.

Image blending is a process in which images are adjusted after the warping process in order to eliminate distortion and discontinuity between the images.

Source image Target image

Superimposed image before blending

Image after blending

Blending is used to obtain high resoluted images such as to view section of tissues.

Most common approach is to acquire several images of parts of tissue at high magnification and assemble them into composite single image which preserves high resolution.

The task of image blending technique is to produce coherent composite image from weighted combination of component image.

Composite images are assembled together to obtain the mosaicised image.

Fig.1: Source image Fig.2: Target image

Fig.3: Mosaiced image

5.APPLICATIONS:

Useful for image stabilization and building high quality images using low-cost imaging equipment's

Remote sensing: monitoring of global land usage,

landscape planning, radar image independent of cloud cover and solar illumination.

Bio-medical imaging: monitoring of healing therapy, monitoring of tumorevolution.Comparison of the patient’s image with digital anatomical atlases, specimen classification.

In video processing.

6.CONCLUSION :

In this project the feature-based matching algorithm is implemented and it has the following stages.

Firstly filtering of images to extract feature points from the images.

Secondly to find matches between sets of points, to get an approximate registration of the 2 images. Hence the same techniques can be used to register PET/CT OR MRI images.

7.REFRENCES:

Inampudi,Ramesh.B. “Imagemmosaicing”, Geoscience and Remote Sensing Symposium Proceedings, 1998. IGARSS '98. 1998 IEEE International 6-10 July 1998.

Satya Prakash Mallick. ”Feature Based Image Mosaicking”.

“PET/CT and MR Image Registration using Normalized Cross Correlation Algorithm and Spatial Transformation Techniques “, B.Balasubramanian , Dr. K. Porkumarann

Image registration and fusion by “Prof.Michael Brady “.

Image compositing and blending by“Prof.Marc Pollefeys” and “Dr.Gabriel Brostow”.

Barbara zitova,jan flusser,dept of Image processing. Institute of information theory and automation. Academy of science Czech republic “ Image registration : a survey”.

Registration of MR image :From2D to 3D,using 8 projection Based Cross Correlation Method J.P.Didon, E Langevin UTC CIMA, 13 Rue de Fonds.

THANK YOU