Wednesday, 29 February 2012

Online Free Learning Resources


 Virtual Lab
  • An Initiative of Ministry of Human Resource Development (MHRD), Goverment of India, under the National Mission on Education through Information Communication Technology.
  • Content Contributing Universities : IIT Delh, IIT Kanpur, IIT Kharagpur, IIT Madras, IIT Roorkee, IIT Guwahati, IIIT hyderabad, Amrita University, NIT Karnataka, College of Engineering, Pune, Dayalbagh University, Agra.
  • COMMENT : I tried for thresholding operation. It worked ok. User interface is not very intuitive. So patient and try to learn it.
  • Website address : http://deploy.virtual-labs.ac.in/labs/cse19/index.php
  • For other subjects visit http://vlab.co.in/

MIT open course ware

Course Description
General Introduction - Estimating Image Motion Field - Pinhole Camera Model - Lambertian Surfaces - Image Motion in 2D -Constant Brightness Equation - Binocular Stereo - Reflectance map: Image Formation, Surface Reflectance, Surface Orientation -Photometric Stereo -Brightness Distribution Function (BRDF) - Shape from Shading -Shape from Gradient - Binary Image Processing -Methods and Applications - The Euler Equations and their Applications - Optical Flow - Motion Vision - Direct Motion Vision - Extended Gaussian Images -Tessellation - Platonic solids - Quaternions - Operations on EGIs - Photogrammetry -Absolute Orientation - Exterior Orientation - Interior Orientation -Relative Orientation
                 University of Berkeley Computer Vision course
Instructor
Jitendra Malik is Arthur J. Chick Professor of Electrical Engineering and Computer Science at UC Berkeley, where he has been on the faculty since 1986. He is also on the faculty of the department of Bioengineering, and in the Cognitive Science and Vision Science groups.Professor Malik is Alumnus of IIT Kanpur. He received his PhD degree from Stanford University in 1985. He is a Fellow of the ACM and IEEE, and a member of the National Academy of Engineering. He has more than 150 research papers and 30 PhD dissertations. He he is one of ISI's Highly Cited Researchers in Engineering.

Computer Vision in simple terms
Computer vision seeks to develop algorithms that replicate capabilities of the human brain - to infer information from the light that fell on the eyes.
  • We can determine distance of the object, object's orientation, relationship with other objects, object's color and texture.
  • We can recognize the object whether it is a chair dog or picture of Dr. Manmohan Singh.
  • We can segment out regions of space corresponding to particular objects and track them over time, such as a cricket fielder chasing a ball.
About the course
In this course face detection, handwritten digit recognition, reconstructing three-dimensional models of cities, automated monitoring of activities, segmenting out organs or tissues in biological images, and sensing for control of robots. They discuss image formation, core image processing operations, and statistical machine learning.

About joining
The classes will start from March 2012. It is free. University of Berkeley will not give any certificate or credit for taking this course. It can be listened at your leisure by downloading the video that will run few minutes. Please register and use the oppurtunity.
  http://www.vision-class.org/

Monday, 9 January 2012

CVIPtools

 A software package for the exploration of computer vision and image processing

One of the primary purposes of the CVIPtools development is to allow students, faculty, and other researchers to explore the power of computer processing of digital images.
Team lead by :: Dr. Scott E Umbaugh
Developed by:
Computer Vision and Image Processing Laboratory
Southern Illinois University @  Edwardsville
Illinois, USA

The new Windows version of CVIPtools, CVIPtools 5.x is implemented in four layers: the algorithms code layer, the Common Object Module (COM) interface layer, the CvipOp layer, and the Graphical User Interface (GUI).

    Brief list of the functionality currently available from the CVIPtools GUI:
          o Edge/Line detection and edge linking
          o Image segmentation - fuzzyc mean,  principal components transform/median cut,
          o Morphological filters -
          o Two-dimensional fast transforms - Fourier (FFT), cosine (DCT),
          o Frequency domain filters - highpass, lowpass, bandpass, bandreject,
          o Feature extraction - binary, RST-invariant, histogram,
          o Feature analysis - Range-Normalization, Unit Vector Normalization,
          o Pattern classification - nearest neighbor, K-nearest neighbor and nearest centroid.
          o Spatial domain image restoration - order filters, mean filters, adaptive filters.
          o Frequency domain image restoration filters - inverse, geometric mean,
          o Noise functions - negative exponential, rayleigh, gaussian,
          o Geometry transformation - user specified mesh. Nearest neighbor,
          o Histogram-based image enhancement - histogram equalization,
          o Pseudo-color enhancement - frequency domain mapping, gray level mapping,
          o Spatial and frequency domain image smoothing and sharpening.
          o Image compression algorithms - bitplane run-length-coding,
          o Support for common image formats - Sun Raster, IRIX, GIF, TIFF, JPEG,
          o Image geometry operations - copy-paste, translate, resize, rotate, zoom.
          o Utilities for arithmetic and logic operations, image file conversion,

CVIPTools based book:

Digital Image Processing and Analysis: Human and Computer Vision Applications with CVIPtools, Second Edition.
ISBN-10: 143980205X   ISBN-13: 978-1439802052 
Publication Year: 2010
CRC Press
Author:: Dr. Scott E Umbaugh
cost :: around 100$

I am not satisfied with the GUI of the tool. It is not very intuitive. I have not downloaded and used. Photo of the BOOK takes 2.5MB. Very few full text papers are there to download. 

Saturday, 7 January 2012

Computational Photography

 Courtesy:: Economist

The principle of focusing rays through an aperture onto a two-dimensional surface remains the same from 5th century BC Camera obscura to latest digital camera.

Computational photography, a subdiscipline of computer graphics, conjures up images rather than simply capturing them.

The best example of computational photography is high-dynamic-range (HDR) imaging. It combines multiple photos shot in rapid succession, and at different exposures, into one single picture of superior quality. Apple added HDR as an option in the iPhone 4.

 HDR idea was developed by  Marc Levoy of Stanford University, with his colleague Pat Hanrahan in 1996. They wrote a journal paper that  described a way of simplifying ligh field mathematically. It is now a reality because "You are getting more computing power per pixel."

 Dr Levoy developed SynthCam and Frankencameras that can improve photos taken in low-light conditions. On June 22nd Ren Ng, a former student of his at Stanford, launched a new company called Lytro,  which promises to launch an affordable snapshot camera.

He used an approach is known as light-field photography, and Lytro's camera will be its first commercial exploration. In physics, a light field describes the direction of all the idealised light rays passing through an area. In Dr Ng's camera light field is created using an array of microlenses inserted in between an ordinary camera lens and the image sensor. By calculating the path between the lens and the sensor, the precise direction of a light ray can be reconstructed.

This ray tracing concept also derived from computer graphics. In computer graphics , the technique is used to paint realistic reflections of one artificial object on another, among other things.

for detailed and more information go to :: http://www.economist.com/node/21522976

Friday, 6 January 2012

DIA in Agriculture

Automated Rice Leaf Disease Detection Using Color Image Analysis

By R.A.D.L. Pugoy and V.Y. Mariano, in  3rd International Conference on Digital Image Processing (ICDIP 2011), Chengdu, China, 15-17 April 2011

Abstract
In rice-related institutions such as the International Rice Research Institute, assessing the health condition of a rice plant through its leaves, which is usually done as manual eyeball exercise, is important to come up with good nutrient and disease management strategies.
In this paper, an automated system that can detect diseases present in a rice leaf using color image analysis is presented. In the system, the outlier region is first obtained from a rice leaf image to be tested using histogram intersection between the test and healthy rice leaf images. Upon obtaining the outlier, it is then subjected to a threshold-based K-means clustering algorithm to group related regions into clusters. Then these clusters are subjected to further analysis to finally determine the suspected diseases of the rice leaf.

Mariano Details
Vladimir Y. Mariano,
Associate Professor, Institute of Computer Science,
The University of the Philippines Los BaƱos.

 
Source:: http://www.ics.uplb.edu.ph/node/477

Document image analysis and recognition


By Simone Marinai 
Document image analysis and recognition (DIAR) is a research field that has its roots in the first Optical Character Recognition (OCR) systems, applied for reading numeric check codes. Nowadays, the technology related to DIAR is used in a broad range of applications, where some information has to be extracted from structured documents existing in different media. Typical applications include, among the others, handwritten character recognition, processing of textual web images, and information extraction from digital libraries. In the digital library community a lot of efforts have been devoted to the digitization of paper collections in order to archive them as document image collections. Large digital archives hare currently available, however their full fruition can be achieved only by accessing the information that is embedded in the digital image. The simple application of Optical Character Recognition (OCR) packages can only partially solve these problems, both for the difficulty of obtaining clean converted text and for the lack of structural description of the document. To tackle this problems either layout analysis methods or document image retrieval approaches can be considered.

Scanning and storage
  • Raw storage
  • Image compression
  • Document image compression
Image pre-processing
  • Noise removal
  • Skew detection
  • Connected components computation
Layout analysis
  • Geometric and logic layout analysis
  • Region segmentation and labeling
  • Page classification
OCR and handwriting recognition
  • Character segmentation
  • Word recognition
Document image retrieval
  • Processing of converted text
  • Retrieval by layout similarity