DIGITAL IMAGE QUALITY IMPROVEMENT APPLICATION FOR TEACHERS AND STUDENTS OF IMAGE PROCESSING SUBJECT

Transcription

1 DIGITAL IMAGE QUALITY IMPROVEMENT APPLICATION FOR TEACHERS AND STUDENTS OF IMAGE PROCESSING SUBJECT Mesterjon 1, Leni Natalia Zulita 2, Yupianti 3,* 1,2,3 Lecturer of Information Technology Study Program, Dehasen University of Indonesia Abstract The background of this research is that there is necessity of reference in form of digital image processing application that can be used directly either by using mobile phone or desktop computer by teachers and students in theaching process of Image Processing subject. One of The purposes of this research is to make teaching application for Digital Image Processing subject. The methods used in designing the application are fuzzy image, filtering and sharpening, and to make the application the researcher uses Matlab application program and does coding by using M-File. The result of this research is an application that can be used by teachers and students; from the application students also get information of the result of digital image quality improvement in various conditions and functions. From this application, digital image processing using matlab and coding using M-File can be figured out also. After doing all digital image processing, a reduced-size file will be gotten. Key words : Digital Image Application, fuzzy image, filtering, sharpening 1. Introduction The advance of information technology has reached all aspects of human life. It brings significant effects. One of them is in digital image enhancement. Though an image is rich with information, sometimes it is degraded in term of quality because of a defect (noise). Noise is a dot in an image that actually is not part of the image, but it is mixed in the image by a cause. Noise is usually because of reckless safekeeping or because of transmission channel (data transferring) while image is produced. In order that image or picture with noise can be interpreted whether by human or machine, the quality of the image needs enhancement. Image enhancement is one of initial processes in image processing. To enhance image quality, the methods used in this research are fuzzy image filtering and sharpening. Fuzzy image sharpening is aimed to sharpen the edge of object in an image. The sharpening operation is done by filtering image through high-pas filter. High-pas filter will filter (or strengthen) high -frequency component (such as edge of an object) and will decrease low-frequency component. As a result, the edge of object appears to be sharper compare to the surrounding. While fuzzy image filtering, is aimed to minimize noise in an image. This research produces digital-image-processing application that can be used as teaching media in junior high school; with goal to enhance digital image quality by using fuzzy image filtering and sharpening. 2. Literature Review and Method a. Literatur Review Image is a representation (illustration), resemblance, or imitation of an object. Image as an output of a data recording system can be optical in the form of photo, analog in the form of video signals such as picture on television monitor, or digital that can directly be saved in a storage device. Digital image as function is two dimensional light intensity f (x,y), where x and y show spatial coordinate, while the value of f at a point (x,y) is comparable with brightness (gray level) of the image in that point. (Gonzales et al., 1913) ISBN:

2 Digital Image Generally, digital image process refers to processing of two dimension image using computer. In a more extend context; digital image processing refers to processing of every two-dimension data. Digital image is an array that contains real or complex values, which are represented with row of specific bit. An image can be defined as function f (x,y) sized M row and N column, with x and y are spatial coordinates and intensity f at coordinate point (x,y) is called intensity or gray level of image at that point. (Kenneth R et al., 1996) Pic. 1. Illustration of Digitalized Image From the above picture it can be explained that Digital image is image f (x,y) where spatial coordinate discretization (sampling) and grey level discretization (quantization) are done. Digital image is a matrix where its row and column index represent a point in the image and its matrix element (that is called as image/pi xel/picture element) represents the grey level at that point. (Kohonen et al., 1996) Image Enhancement is a process to get image that is easier to be interpreted by human eyes. In this process, the specific characteristics that exist in an image are emphasized to appear. Quality enhancement is needed because it is often the image that has been the object of discussion has bad quality, for example, the image is too bright/dark, less sharp, blur, etc.(panjaitan et al, 2007) Image Filtering, Filtering process on image processing is commonly called filtering. In this process, the new pixel value is commonly counted based on neighbor s pixel. For example, low pas filter means filtering low frequency component. Low pass filter produces blur/soft image.(abdul et al., 2006). All coefficients of low pass filter are positive numbers. Image filtering is aimed to minimize noise in an image. Noise usually appears as a result of bad sampling (sensor noise, photographic grain noise) or because of transmission channel (in data transmission). Noise in image is commonly in the form of intensity variation of a pixel that does not correlated with neighbors pixels. (Gasimov et al., 2002). Visually, noise is easily seen by eyes because it appears to be different from its neighbors pixel. Image Sharpening, Image sharpening is aimed to sharpen the edge of an object in an image. Image sharpening is done by filtering image in high-pass filter. (Jantzen et al., 1998). Because image sharpening affects more on edge of object, it is called edge sharpening or edge enhancement. The edge of object appears to be sharper compare to the surrounding. Image sharpening is aimed to sharpen edge of object in image. Image sharpening is contrary to image softening/ filtering because this operation omits the soft part of the image. Image sharpening is done by filtering image through high-pass filter. High pass filter will filter (strengthen) high frequency component (for example edge of object) and will decrease low frequency component. Thus, edge of object appears sharper compare to the surrounding (Bellman et al., 1997). Fuzzy, fuzzy initially was from mathematic and system theory of L.A.Zadeh [9], in Literally, fuzzy means not clear/blur, uncertain. Fuzzy unity is branch of oldest mathematic, which studies the process of random number: probability theory, mathematical ISBN:

3 statistic, in formation theory etc. Problem solving using fuzzy unity is easier than by using probability theory (measurement concept) (Klir et al., 1995). Fuzzy Image Processing, fuzzy image processing is collection of all approaches that comprehend, present and process image, motive and segment as fuzzy determination. Presentation and process depend on fuzzy determination technique and problem to solve [11]. Fuzzy image processing has 3 mains stages: image fuzzification, modification of member value, and image defuzzification (Kusuma Dewi et al, 2004). b. Research Method The method used in solving problem to create digital image enhancement application using fuzzy image filtering and sharpening is prototyping model (Whiten et al., 2004). This method is selected in order that the built application can run well. Besides, the main reason of taking this method is that there is repetition of data taking after evaluation or testing that is still less accurate. For example, if the testing result of a prototype does not fit with the goal, information collection is to be done again to complete what it is missing. Thus, this method is very suitable in creating this application. (White, et al., ) The illustration of prototyping model is seen in picture 1 Pic.2. Prototyping Model Stages in prototyping model can be explained as follows: a. Information collection is aimed to take data and information which is needed in wouldbe built application. b. Stage of building/repairing prototype is done by elaborating the result of information collection and also repairing the prototype that has been built. c. The next stage is to test the prototype result, which is the testing of the result of the prototype that has been built. This is done to minimize unexpected errors. d. These processes are repeated if the result is not satisfactory yet. If the result has already been satisfactory this process will stop at the stage of prototype testing. 3. Research Result This research results image enhancement application using fuzzy image filtering and sharpening using GUI Matlab R2013a based on computation with Matlab programming language. The application is aimed for user and data processor of digital image. This application is operated on windows platform devices that use array concept as standard of library variable of Window GUI Matlab. Thus, the application integrates computation ability, visualization, and programming in a single and user friendly environment. ISBN:

4 Pic. 3. Display of Application Start Menu Picture 3 shows display of start menu that is used to take image file, the taken image is the original one. To take the wanted image click Open file button, so that sub menu of existing image file in the data base appears. Pic. 4. Image file Picture 4 shows interface of digital image that is going to be selected, the file that is going to be used should never been edited. After selecting the image immediately click on the image. Pic. 5. Selected image file Picture 5 shows after image is selected, the metadata of the image will be shown, sized in x,y of the image and the image capacity sized in Kb. After that, filtering can be done to enhance image quality by clicking the process button in the Filtering dialog and will be shown in the box. Besides, the resolution and size of the file after filtering will be shown. ISBN:

5 Pic. 6. Filtering Process Picture 6 shows filtering process. After filtering is done, the image that has been processed will be shown with its metadata, which contains x,y of image size and the size of the image in kilo bit. Later, image in the form of Bitmap file can be stored in data base by clicking save button. Pic. 7. Selected image file Picture 5 shows the process of sharpening by clicking on process button in sharpening dialog and the result will be shown in the box and is saved in the data base. a. Testing Result Testing result with some samples can be seen in the following table: Table 1. Testing Result No First image Original file Test results Filtering Test results Sharpening File size Citra 1 File size Citra 2 File size 149,265 Kb 21, 943 Kb 17, 93 Kb 129,265 Kb 20, 987 Kb 12, 983 Kb 107 Kb 18 Kb 10,23 Kb 99,295 Kb 18, 943 Kb 10, 93 Kb 179,205 Kb 26, 43 Kb 22, 993 Kb Kb Kb Kb Kb Kb Kb ISBN:

6 Kb Kb 9.77 Kb Kb Kb Kb Kb Kb Kb After testing is done with 10 image sample as above, it is found that image enhancement using image filtering and sharpening can be used to soften and enhance image quality. Besides, image size is smaller than the original. From the testing result it can be concluded that sharpening process has better result in image enhancement. References Gonzales, Rafael C. and Woods, Richard Digital Image Processing.USA: Addison- Wesley Publishing Company. Kenneth R. Castleman.1996.Digital Image Processing.Prentice Hall. Kohonen, T., Hynninen, J., Kangas, J., Laaksonen, J.1996.SOM_PAK: The selforganizing map program package. Report A31.Helsinki University of Technology, Laboratory of Computer and Information Science, Espoo, Finland. Also available in the Internet at the address Pandjaitan, Lanny W Dasar-dasar Komputasi Cerdas.Yogyakarta:Andi Offset. Siang, Jong Jek, Drs., MSc.2005.Jaringan Syaraf Tiruan & Pemrogramannya Menggunakan MATLAB.Yogyakarta:Andi. Abdul Halim Fathoni, Artikel Metode Horisontal - Bahasa lmatematika, Gasimov, R. N., Yenilmez, K., Solving fuzzy linear programming problems with linear membership functions, Turk J Math. 26, , Jantzen Jan, Tutorial On Fuzzy Logic, Technical University of Denmark, Department of Automation, Bellman, R. and Zadeh, L.A., Decision making in a fuzzy environment, Management Science,17, , 1970 Klir, G.J., Yuan, B., Fuzzy Sets and Fuzzy Logic-Theory and Applications, Prentice-Hall Inc., 574p., Klir, G.J., Folger T. A., Fuzzy Sets, Uncertainty and Information, Prentice Hall International, Inc., Kusumadewi S., Purnomo H., Aplikasi Logika Fuzzy, Untuk pendukung keputusan, Graha Ilmu, Lai, Y. J. and Hwang, C. L., Interactive fuzzy linear programming, Fuzzy Sets and Systems, 45, , Lai, Y. J. and Hwang, C. L., Fuzzy Multiple Objective Decision Making: Methods and Applications, Springer-Verlag, New York, Whiten L. Betley D. Diiman C.. System analysis and desingn methods: Metode desain dan analisis sistem edisi6.andi opset :yogyakarta Lippeveld T, R. Sauerborn R,C Bodart. Design and Implementation of Health Information System. Geneva : World Health Organization Popescu, C. and Sudradjat, S., Parameter estimation for fuzzy sets, IJPAM, accepted Novembers 6, ISBN:

7 Popescu, C., Sudradjat, S. and M. Ghica, On least squares approach in a fuzzy setting, Conferinţă a Societăţii Probabilităţii şi Statistică din România, Aprilie Puri, M.L., Ralescu, D.A., Fuzzy random variables, J. Math. Anal. Appl. 114, Sciences, 15, 1-29, Rommelfanger, H., Fuzzy linear programming and applications, Europan Journal of Operational Research, 92, , AUTHORS First Author- Dr.Mesterjon,S.Kom,M.Kom, is Lecturer of Information Technology Study Program, Dehasen University of Indonesia, PH mesterup@yahoo.co.id Leni Natalia Zulita, S.Kom, M.Kom, is Lecturer of Information Technology Study Program, Dehasen University of Indonesia, PH natalia.af25@gmail.com, Yupianti, S.Kom, M.Kom, is Lecturer of Information Technology Study Program, Dehasen University of Indonesia, PH Yupiantiprana@gmail.com ISBN: