CURRENCY DETECTION AND DENOMINATION SYSTEM USING IMAGE PROCESSING Pranjal Ambre 1, Ahamadraja Mansuri 2, Harsh Patel 3, Assistant Prof. Sunita Naik 4 B.E. Computer Engineering, VIVA Institute of Technology, Mumbai University Pranjalambre40@gmail.com Ahamadmansuri47@gmail.com Harsh2521@gmail.com sunitanaik@viva-technology.org Abstract Counterfeit notes are one of the biggest problem occurring in cash transactions. It is easily possible for a person to print fake notes because of advances in printing, scanning technologies and use of latest hardware tools. Detecting fake notes manually becomes time consuming and is an unclean process; hence there is a need of automation techniques with which currency recognition process can be efficiently done. Many techniques have been proposed with the use of MATLAB feature extraction with HSV color space and other applications of image processing. Also it is very difficult to count different denomination notes in a bunch. Some filters are applied to extract denomination value of note. The system uses different pixel levels in different denomination and Neural Networks matcher techniques. In this paper the system is formulated to implement fake note detection and denomination identification using MATLAB algorithm for Rupee, Dollar and Yen and Pound currencies. Keywords - Currency, Detection, Denomination, Image Processing, MATLAB I. INTRODUCTION Any setup or system can easily recognize a face, understand spoken words, read characters and identify car keys in pocket by feel belies the astoundingly complex processes that underlie these acts of pattern recognition. For multiple currency detection and denomination system is an image processing technology that is used to detect and identify currency amount. Identifying multiple currencies in a bunch is quite tedious and time consuming process.this system can help human in order to live a better life.the proposed system communicates with web cam, catches video frames which include a visible image of currency amount and processes them. Various methodologies are used on the image. The selected area of the image is processed and analyzed with their parameters. Once the image of the currency amount was detected, its digit is recognized it will display on the user interface.this system will be developed using MATLAB. MATLAB is a high performance language for technical computing. II. METHODOLOGY There are various algorithms which work for single currency. But this paper gives the detection and denomination of multiple currencies. The steps are as follows:- A. LUV color space transformation The first step is transforming JPEG or PNG image of the currency to XYZ transformation. The XYZ transformation is converted to LUV and then the mean, color variance and skewness of each channel (L, U and V) is calculated. This is done by using the Colour_luv Function as shown in the Fig.1.
Fig.1 Color luv function fet B. Edge Directional Histogram 1) Edge directional histogram: The basic idea in this step is to build a histogram with the directions of the gradients of the edges (borders or contours). The convolution against each of this mask produces a matrix of the same size of the original image indicating the gradient (strength) of the edge in any particular direction. It is possible to count the max gradient in the final 5 matrix and use that to complete a histogram. 1x5 edge orientation histogram is computed (horizontal, vertical, 2 diagonals and 1 non-directional) 2) Input Image should be a RGB Image: An RGB image, sometimes referred to as a true color image, is stored as an m-by-n-by-3 data array that defines red, green, and blue color components for each individual pixel. RGB images do not use a palette. The color of each pixel is determined by the combination of the red, green, and blue intensities stored in each color plane at the pixel's location. Graphics file formats store RGB images as 24-bit images, where the red, green, and blue components are 8 bits each. This yields a potential of 16 million colors. The precision with which a real-life image can be replicated has led to the nickname "true color image." C. Total Feature Extraction It consists of three feature color, edge and texture feature. In this we are calling the respective functions to store all the features in function fet as shown in the Fig.2. 1) Color Feature: Let denote an image and α be a pixel in I. The color feature extraction is to define a function F: I Q where Q is the set of representative colors, such that F maps a pixel α to representative color. The extracted color feature can be represented as a color histogram. 2) Edge Feature: The shape of an object refers to its physical structure and profile. Shape features are mostly used for finding and matching shapes, recognizing objects or making measurement of shapes. Moment, perimeter, area and orientation are some of the characteristics used for shape feature extraction technique. The shape of an object is determined by its external boundary abstracting from other properties such as color, content and material composition, as well as from the object's other spatial properties. 3) Texture Feature: This method based feature extraction is that it provides a good symbolic description of the image; however, this feature is more useful for image synthesis than analysis tasks. This method is not appropriate for natural textures because of the variability of micro-texture and macro-texture.
Fig.2 Total feature extraction III. Design Flow of Proposed system Fig 3. Flow chart of the proposed system The steps for the above Fig.3 are explained as follows: A. Image Acquisition Image acquisition in image processing can be broadly defined as the action of retrieving an image from some source, usually a hardware-based source, so it can be passed through whatever processes need to occur afterward. Performing image acquisition in image processing is always the first step in the workflow sequence because, without an image, no processing is possible. In this step the image that is acquired is completely unprocessed and is the result of whatever hardware was used to generate it, which can be very important in some fields to have a consistent baseline from which to work. One of the ultimate goals of this process is to have a source of input that operates within such controlled and measured guidelines that the same image can, if necessary, be nearly perfectly reproduced under the same conditions so anomalous factors are easier to locate and eliminate. B. Gray Scale Conversion The image of the currency is converted into gray scale format:-
To convert an RGB image to grayscale, we have use the RGB2GRAY command from the Image Processing Toolbox. The second method is we can use the standard NTSC conversion formula that is used for calculating the effective luminance of a pixel: Intensity = 0.2989*red + 0.5870*green + 0.1140*blue (1) C. Edge Detection The edge detection of the note is carried out to check the size and to test the corners for its sharpness. Edge detection is used in image analysis for finding region boundaries. Edge and contours play a dominant role in human vision and probably in many other biological vision systems as well. Not only are edges visually striking, but it is often possible to describe or reconstruct a complete figure from a few key lines. D. Artificial Neural Network (A.N.N) Artificial Neural Networks are used for the forecasting purpose. The first layer is an input layer where external information is received. The last layer is an output layer where the problem result is obtained. E. Image Segmentation After the gray scale conversion of the image, the image is divided into various segments of security features to test the originality of the currency. After that Artificial Neural Network is use for pattern Recognition. F. Characteristics Extraction The important characters on the currency i.e. the features of it are extracted to test the originality. G. Comparison The security features on the currency and that feed into the system are compared. After comparing the security features we get to know the result whether the currency is fake or original. If it is fake then the warning message will be displayed on the screen. If the currency is original then its denomination i.e. the exact value is displayed. IV. EXPECTED RESULTS A. Currency recognition The proposed system recognizes the currency and shows the output with currency name for respective image of currency. 1) Indian currency: Here Rupee is detected as shown in Fig.4.
Fig.4 Indian Rupee Detection 2) Dollar currency: In this Dollar is detected as shown below in Fig.5. Fig.5 Dollar Detection B. Currency Detection The proposed system detects the originality of the currency with its denomination for Rupee, Dollar, Yen and Pound. V. CONCLUSION The drawback in the previous system was that, one system was only recognizing the currency and the second system was just identifying the denomination value of the currency. These both systems were only valid for single currency i.e. Indian currency only. So after considering all the drawback of the above system the further hybrid system will recognizes the fake currency and also determines its correct value. The proposed system is user friendly with simple methods and is reliable. The system has been added with the currency security features its order to determine its originality. At the same time the system is applicable for multiple currencies i.e. Indian, Euro, Dollar. The currencies can be easily detected in a moment. The denomination helps to count the currencies in a bunch easily. VI. REFERENCES [1] Binod Prashad Yadav, C.S.Patil, R.R.Karhe, P.H Patil; An automatic recognition of fake Indian paper currency note IJESIT Volume 3, Issue 4, July 2014.
[2]Shatrughan Modi; Dept. of Computer Science; Thapar university, Dr. Seema Bawa; Dept. of Computer science; Thapar university, Automated Coin recognition system using ANN IJCA Volume 26, July 2011. [3]Vipin Kumar Jain; lecturer; department of computer science; S.S Jain Subodh P G College Jaipur, Dr.Ritu Vijay; Head;department of Electronics; Bhanasthali university; Indian Currency denomination identification using image processing techique IJCSIT, Vol 4, 2013. [4] Hanish Agarwal, Padam Kumar, "Indian currency note denomination recognition in color images", IJACEC, Vol. 1, ISSN 2278 5140. [5] Krishan, G.2010.Image Processing Based Feature Extraction of Indian Currency Notes. PhD diss., Thapar University [6] Sanjana, Manoj Diwakar, Anand Sharma, "An Automated recognition of Fake or an currency notes in Machine vision", IJCSMS,Vol. 12, April 2012. [7] Sannakki S. S, and Gunjale, P. J 2014, Recognition and Classification of Currency Notes using Discrete Wavelet Transform, International Journal of Emerging Technology and Advanced Engineering, Vol.4, Iss [8]J. Smokelin. Wavelet feature extraction for image pattern recognition. In Proc. SPIE, volume 2751, 1996, pp 110-121. [9] F. Takeda and S. Omatu. Image Processing and Pattern Recognition, Academic Press, 1998, pp 133-160. [10] E. Choia, J. Lee, and J. Yooni. Feature extraction for banknote classification using wavelet transform, In Proc. 18th International Conference on Pattern Recognition, 2006, pp934-937. [11] F. Takeda and T. Nishikage, Multiple kinds of paper currency recognition using neural network and application for euro currency. In Proc. IEEE International Joint Conference on Neural Networks, 2000, pp 143-147. [12] D. A. K. S. Gunaratna, N.D. Kodikara and H. L. Premaratne, ANN Based Currency Recognition System using Compressed Gray Scale and Application for Sri Lankan Currency Notes-SLCRec, Proceedings of world academy of science, engineering and technology, vol. 35, ISSN 2070-3740, Nov 2008. [13] Kalyan Kumar Debnath, Sultan Uddin Ahmed, Md. Shahjahan, A Paper Currency Recognition System Using Negatively Correlated Neural Network Ensemble, Journal of Multimedia, vol. 5, No. 6, December 2010. [14] T Ojala,M Pietikainen, and D Harwood, A comparitive study of texture measures with classification based on feature distributions, Pattern recognition, Vol.29, No. 1, 1996,pp.51-59. [15] H. Hassanpour, A. Yaseri, G. Ardeshiri, Feature Extraction for Paper Currency Recognition, IEEE Transactions,1-4244-0779-6/07,2007.