The ICE Tool Feng Wen Qi Yuan Kin Wah Leung
Presentation Overview Project goal Interactive GUI Introduce image enhancement techniques Integration with Matlab™ Implementation of image enhancement techniques Potential advancement of ICE tool
ICE Tool What is ICE tool? ICE = Image Contrast Enhancement Capable of executing various image enhancement techniques Provides easy to use interface Can be altered according to desire needs if necessary
Project Goal To implement an interactive GUI capable of enhancing images Research image enhancement techniques Programming an interactive GUI Integrating with Matlab™ libraries Implementing image enhancement functions make sure functions performed correctly
Interactive GUI The Interactive GUI (graphic user interface) As user friendly as possible Created using Java™ - A programming language from Sun Microsystems - Provides great system portability Created as a Java™ frame application GUI features Ability to load and save desired images Displays original and modified image on the same panel Easy menu bar browsing
GUI Features Ability to load and save image Ability to display both original and modified image on same screen
GUI Features (cont…) Easy toolbar browsing Combines simple image enhancement methods
Image Enhancement Techniques Contrast Enhancement Histogram equalization - Image quality can be improved by altering histogram - Calculates the ideal transformation from the histogram of the image - All gray levels used has a tendency to enhance image contrast Transformation Function: T(f ) can be calculated from the following relation:
Image Enhancement Techniques (cont…) Noise Removal Filter – removes dots or speckles on image (equivalent of low-pass filtering) Average Filter (Mean) - Replace each pixel by the average of the window area pixels - Has the effect of smoothing image - Larger window size removes noise more effectively while - At the expense of blurring the details Median Filter - Replace each pixel by the median of the window area pixels - More effective against impulse noise (aka salt and pepper) - Can retain details and edges better than averaging filter
Image Enhancement Techniques (cont…) Deblurring Wiener Deblurring - Generalized inverse filter - Effective when information regarding frequency characteristics are known, at least to a degree Lucy-Richardson - Effective when the PSF (point-spread function) is know but little information is available for the noise Sharpening Enhances details and edges Line structures can be obtain by applying high-pass filter
Integration With Matlab™ Benefits Allows the access of the large Matlab™ function library - The Matlab™ math function library - The Matlab™ image processing function library Integration process Use of an software engine to link Matlab ™ and Java ™ GUI together Implement the functions to the appropriate buttons
Integration With Matlab™ Incorporate with JMatLink A Java engine capable of linking Java ™ applications and Matlab ™ - Use of native methods, no source code need to be changed - Created by Stefan Muller Edit autoexec.bat to set path to Matlab ™ and Java ™
Research Matlab ™ code Must know the codes for executing all of the image enhancement techniques Ex: for histogram equalization I = imread(‘abc.jpg’); J = histeq( I ); Image Enhancement Implementation Implement the code to the Java ™ interface buttons Every component assigned the appropriate Matlab ™ code Press of buttons send Java ™ code to Matlab ™ for execution
Image Enhancement Implementation (cont…)
Summary Successfully creating an functional interactive GUI using Java Java was integrated with Matlab™ through JMatLink The Matlab™ code was associated with every button in GUI Additional features and improvements can be made
Future Advancement of ICE Tool Try to make it a standalone application without Matlab™ Addition of more image enhancement techniques Addition of more features such as help documentation, zoom, etc Package into an easy to install application