GISdevelopment.net ---> Education ---> Papers / Articles




Digital Image Processing Simulator (DIPS) - A Software for Teaching Digital Image Processing Concepts

Chinmaya S. Rathore
Associate Professor, Applied Computer Technology
Indian Institute of Forest Management, Bhopal, 462003, India

Introduction
Digital Image Processing (DIP) is taught as part of a number of remote sensing courses around the world. In a large part, DIP is essentially a software centric subject which requires a good conceptual understanding of image processing algorithms. Due to the nature of the subject, the teaching requirements of digital image processing are usually quite demanding. A number of concepts in digital image processing are relatively difficult to communicate using static mediums such as transparency sheets and PowerPoint presentations. For example it may not be the easiest of things to explain to a fresh group of students using diagrams, how a digital convolution filter works, how it performs its math and how the filter kernel moves to the next column or row. Concepts like edge enhancements, ratios, classification and accuracy assessments likewise can pose problems of proper demonstration for the teacher due to the intricate nature of calculations involved in these operations.

The usual cycle for remote sensing teaching involves teaching of concepts in the class followed by demonstrations on the image processing system. This cycle has a major learning gap. While algorithmic explanations are taught in the classroom, the image processing system does not give the student a live, 'under the hood' view of 'how an operation has been completed. The image processing system just simply completes an operation. Even on the smallest of images (like 512 by 512 pixels), it may be very difficult for the student to relate how the image processing system is internally working on each image pixel while implementing an image processing algorithm. The result is that the student cannot usually 'see' the calculations explained in the class when working on an image processing system. The transition from the class to the image processing system may therefore be without the requisite theoretical background.

The Digital Image Processing Simulator or DIPS is essentially a Digital Image Processing learning tool that tries to bridge this learning gap between the class and the image processing system. It has been designed in a way that it can be shared between the student and the teacher. It can be used by the teacher to conveniently explain difficult to understand fundamental image processing concepts during a lecture, while at the same time, it can also be used by the student to follow up concepts explained in the class. It is felt that the use of this software can help create a more helpful learning environment, which can facilitate concept assimilation and knowledge transference.

Development Parameters
DIPS has been written in Visual Basic and in its current zip version for download is around 290 Kb in size. It works on the Windows platform and requires VB6 Runtime file for its operation, which can be downloaded free from a number of sites the Internet. No installation is needed. DIPS has evolved purely on personal interest and effort. This work was not funded and was pursued by the author on his own accord. The conceptualization of this software involved visualizing core image processing concepts and the presentation of these concepts in an easy to understand manner. While conceptualizing this software, the author set some important design constraints. These were:
  • The software should be small in size to permit easy transmission.
  • The software should show 'under-the-hood' calculations of all concepts.
  • The software should provide interactivity to explore a concept by permitting them to change things and see the effect of these changes.
  • The software should cover a core body of image processing concepts.
  • The software interface should be intuitive and easy to understand.
  • It should have learning pointers and exercises that expose concepts from different angles.
In writing this software, the author had to avoid the use of special VB controls that would add to the size of the software. In pursuance of the size constraint a simple text file was used to improvise a basic help system. The default data and exercises in various modules were designed to expose different facets of the image-processing concept under study. The screen presentation of each concept covered by DIPS was conceived in a manner such that the concept can be graphically communicated in detail with ease. In addition, screens had to be designed such that calculations for each concept could be seen in a step by step mode.

Concept Coverage
DIPS is primarily aimed at introductory digital image processing courses in remote sensing. DIPS should be of use to University/ College teachers , students undergoing Remote Sensing Courses, and other training organizations. DIPS covers a core body of digital image processing concepts. DIPS has been specifically designed to expose detailed calculations of image processing algorithms using small sample images. The users can change these small sample images to build their own examples and add to the pool of exercises. DIPS covers the following important image processing concepts:
  • Additive and Subtractive Primary Colours
  • Display of Colour on Multi-band Digital Images
  • Gun Band Assignments
  • Linear Contrast Stretching
  • BSQ,BIL and BIP Digital Data formats
  • Line dropout and Salt and Pepper noise removal
  • Operation of Low Pass and high pass convolution filters
  • Image Ratioing
  • Normalized Vegetation Difference Index
  • Image Differencing (subtraction)
  • Linear Edge Enhancement
  • Density Slicing
  • Supervised Classification
    • Signature Definition
    • Minimum Distance to Mean Classifier
    • Parallelepiped Classifier
    • Comparison of Classification Outputs
  • Accuracy Assessment
    • Random Selection of Pixels on Classified and Ground maps
    • Computation of Error Matrix and % Correct
    • Omission and Commission Errors
    • Computation of Kappa Statistic
Plates 1-4 show selected screenshots of some modules of the software.

Availability
DIPS is available as a free download from the DIPS website at the following URLS http://www.iifm.org/downloads/dips

The above site is also mirrored at http://210.212.144.2/dips/index.html
The above websites have all requisite links and help files needed to run DIPS at one convenient location.


Plate 1 : The Main menu of DIPS



Plate 2 : Convolution Filter Explorer permitting the user to see the operation of High and Low Pass Filters



Plate 3 : Permits the User to follow the process of supervised classification on sample data.



Plate 4: Accuracy Assessment Module permits the user to learn about the accuracy assessment calculations

© GISdevelopment.net. All rights reserved.