The Grading Multiple Choice Tests System via Mobile Phone using Image Processing Technique

— Grading devices are expensive causing budget waste, in addition some are difficult to use. Therefore, an objective test grading system via Android mobile phone was developed to save cost and time in grading. The system uses image processing technique developed by Java. A camera on a mobile phone was used to capture the edge of answers and an equation of geometric simulation of digital camera sensor was applied to identify answers selected from calculation of pixel intensity in real time. The objective test grading system via Android mobile phone can work effectively and accurately more than 95%.


Introduction
Nowadays, smartphone technology plays an important role in daily life. There are plenty applications available to serve users' needs. Technology has become an essential part in people's daily life and studies on technology has been key to educational driving. Grading is one of the systems that applies technology. Grading devices are expensive which cause budget waste. It also takes more time when using the device at the same time. Therefore, an objective test grading system via Android mobile phone using image processing technology was developed to save cost and time in grading.
Image processing theory was applied to objective test grading in several means including an automatic marking system for multiple choice test on network [1], an imageprocessing oriented multiple-choice checking system [2] and a flexible and cost-effective multiple-choice exam system [3]. The systems use a scanner to receive images into a computer. Then, image files are processed. Multiple-choice exam program with parallel processing [4] is a system development that enables a grading device to work faster. Additionally, an investigation of the color and density of different multiplechoice answer sheet input factors [5] using image processing techniques is a system development that makes grading faster. In order to develop objective test grading system, the researcher developed an objective test grading system via Android mobile phone to make grading faster, more convenient and to save cost and time in grading.

Color Retention in Matrix
It is a part of image processing preparation. A colour pixel positioned in this digital image is a pixel, referring to brightness level of the digital image in a two -dimensional M x N matrix as shown in

Grayscale
Adjusting grayscale before processing an image makes image processing easier. Grayscale adjustment is shown in

Edge Detection
Edge detection identifies edges of an object in the image which is outstanding and important to knowing that object captured by a computer. Edge detection identifies outstanding features of the object. Outstanding feature that people could see is a ridge or an edge of the object. When the light reflects on the object, this part is brighter than other areas. In general, good features of edge are thin (an edge of the image should have only one width), consistent (edges within the same image should be consistent) and a position of a pixel should be accurate. The edge of the accurate image is within an area where intensity of the pixel is higher. The edges are represented through different numbers. The thick edges of the image (2 points of width) and inconsistent edges of the image are shown in Fig 3. (a) The thick edges (2 points of width) (b) Inconsistent edges

Image Processing Algorithm
Logical reasons and mathematics are applied to choose the next methods or procedures. The final method is to subdivide and sort operational procedures in image processing in order to increase effectiveness, rapidness and accurate results. Mathematical principles used in processing are shown in Fig 4, 5, 6. Symbols representing a distance between P and Q is PQ.

System Design
A user uses the system via Android mobile phone, using a camera on the phone to capture an image for processing. The system consists of 2 main parts including an interface and processing. The interface is divided into 4 parts including keeping answer key, checking answers, value clearing and existing the program as shown in

Image Processing Design
The system processes the image received according to image processing, consisting of 4 main steps including importing input, processing preparation, image processing and result displaying as shown in      The choice values to be considered are kept and the coordinates of cross marks are identified as shown in Figure 16. MAXAVG is an average of pixel intensity which is higher than an average of overall pixel intensity within the square. MINAVG is an average of pixel intensity which is lower than an average of overall pixel intensity within the square compared with CV of each choice. The answer is the choice with the highest CV, when examining other items that have no cross marks or that have more than 1 cross mark which will score 0.
Score result is displayed on a mobile phone screen.

Experimental Results
The accuracy tests for the objective test grading system via Android mobile phone were carried out via 2 means.

4.1
Examining for no cross marks.

4.2
Examining for a single-cross mark in one item. Fig. 18. Examine for a single-cross mark in one item. Of 1600 items, the system yielded 1550 correct items on 1550 items or 96.875 percent. According to the result, examining answer sheets without any cross marks yielded more accurate results than examining answer sheets with more than one cross mark in one item.

Conclusion
The Grading multiple choice tests system via Android mobile phone was developed for actual use in objective test grading, using image processing developed from library of Java. Of 1600 items, the system yielded 1550 correct items or 96.875 percent.
The Grading multiple choice tests system via Android mobile phone is effective and accurate. It is convenient for users, saves time in grading and saves cost of objective test grading devices. However, the developed algorithm can be used with one answer sheet format only.

Authors
Worawut Yimyam received the B.E. degree in Computer engineering from the Rajamangala University of Technology Thanyaburi, Thailand, in 2000, and the M.E. degree in Information Technology from the University of Rangsit, Thailand, in 2007 and the study Ph.D. degrees in Information Technology from Information Technology, King Mongkut's University of Technology North Bangkok, Thailand. He has held instructor positions at faculty of business administration from Rajamangala University of technology thanyaburi, Thailand, in 2001. He is currently a Lecturer at Faculty management science from Rajabhat phetchaburi, Thailand.