Denunciar

Meghana KantharajSeguir

31 de May de 2018•0 gostou•906 visualizações

31 de May de 2018•0 gostou•906 visualizações

Baixar para ler offline

Denunciar

Dados e análise

This presentation explains the process of recognizing handwritten mathematical symbols using Neural Networks in MATLAB.

Meghana KantharajSeguir

Digit recognitionbtandale

Handwritten Digit Recognition(Convolutional Neural Network) PPTRishabhTyagi48

Handwritten Digit Recognition using Convolutional Neural NetworksIRJET Journal

Text detection and recognition from natural sceneshemanthmcqueen

Feature selectiondkpawar

Automatic handwriting recognitionBIJIT GHOSH

- 1. Mathematical Symbol Recognition
- 2. Problem Statement • Recognize symbols that have been input. • Inputs are called test data. • The classifier classifies with the training gained by training data. • The ration of training to test data is 70:30. • The symbols to be recognised are : Minus(-) Plus(+) Multiplication(*) Division( / ) Greater than(>) Lesser than(<) Opening square brackets( [ ) Closing square brackets( ] )
- 3. Tools used. • MATLAB R2011a – Image processing toolbox – Neural network toolbox
- 4. Methods of character recognition • We have applied mathematical character recognition on a dataset using the following methods – Symbol recognition using image pixel intensity – Symbol recognition using image features of zoning.
- 5. Steps involved in mathematical symbols recognition:
- 6. Pre-processing: • The pre-processing is a series of operations performed on the scanned input image. It essentially enhances the image rendering it suitable for segmentation. Pre-processing aims to produce data that are easy for the character recognition systems to operate accurately. • The Operations we performed on the scanned input images are: Binarization :It converts a grey-scale document image into a binary document image. Edge detection: is an image processing technique for finding the boundaries of objects within images. Morphology: is performed using image dilation and image filling.
- 7. 1. Input greyscale Image 2. Binarized Image
- 8. 3. Noise Filtering 4. Edge Detection
- 9. 5. Dilation 6. Fill Image with bounding boxes
- 10. Segmentation: Is an image processing technique for finding the boundaries of objects within images. It works by detecting discontinuities in brightness. Thresholding methods and were used to achieve the same. Blob analysis: this was used to find all the objects on the image, and find the properties of each object. Plotting the Object Location : was used to plot the object locations. Feature extraction uses SobelTechnique for edge detection
- 11. Feature Extraction: • images cropped close to the boundary • feature vector consists of the intensity value of a pixel corresponding to pixel in the binary image • Input is negative of the figure • input range is 0 and 1 (0 - black and 1 – white) • value show the intensity of the relevant pixel • 25 X 25 matrices
- 12. • UNIVERSEOF DISCOURSE – shortest matrix that fits the character skeleton – features extracted from the character image include the positions of different line segments in the character image – every character image should be independent of its Image size • Zoning – image divided into windows of equal size, feature is done on individual windows – image was zoned into 9 equal sized windows – Feature extraction was applied to individual ones rather than the whole image – positions of different line segments in a character skeleton becomes a feature if zoning is used
- 13. • CHARACTERTRAVERSAL – starts after zoning is done on the image – Each zone is individually subjected to the process of extracting line segments – first the starters and intersections in the zone are found and then populated in a list – Minor starters are found along the course of traversal – Algorithm starts by considering the starters list – minor starters obtained so are processed – algorithm starts with the minor starters – All the line segments obtained during this process are stored with the positions of pixels in each line segment – Once all the pixels in the image are visited, the algorithm stops.
- 14. • DISTINGUISHING LINE SEGMENTS After line segments have been extracted from the image they have to be classified into any one of the following line types. • Horizontal line. •Vertical line. • Right diagonal line. • Left diagonal line.
- 15. • FEATURE EXTRACTION After the line type of each segment is determined, feature vector is formed based on this information. Every zone has a feature vector corresponding to it. Under the algorithm proposed, every zone has a feature vector with a length of 8.The contents of each zone feature vector are 1) Number of horizontal lines. 2) Number of vertical lines. 3) Number of Right diagonal lines. 4) Number of Left diagonal lines. 5) Normalized Length of all horizontal lines. 6) Normalized Length of all vertical lines. 7) Normalized Length of all right diagonal lines. 8) Normalized Length of all left diagonal lines. 9) NormalizedArea of the Skeleton.
- 16. Classification: • This was achieved by using neural networks , which was a feed forward backward propagation mechanism. • The neural network consists of 2 layers with 65 neurons in the input layer and 8 neurons in the output layer for FEATURE EXTRACTION with pixel intensity • The neural network consists of 2 layers with 85 neurons in the input layer and 8 neurons in the output layer for FEATURE EXTRACTION with zoning
- 17. Recognition: • The input data was divided for the following – Training ( 70% ) – Validation ( 5% ) – Testing ( 25% )
- 21. Output:
- 22. Comparision of results • Using pixel intensity feature extraction technique We get an efficiency of 96% • Using zoning feature extraction technique we get an efficiency of 83%
- 23. References • Chirag I Patel, Ripal Patel, Palak Patel, “Handwritten Character Recognition Using Neural Networks”, International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011. • Kauleshwar Prasad, Devvrat C Nigam, Ashmika Lakhotiya, Dheeren Umre, “Character Recognition Using Matlab’s Neural Toolbox”, International Journal of u- and e- Service, Science andTechnologyVol. 6, No. 1, February, 2013. • Ashutosh Aggarwal, Rajneesh Rani, Renu Dhir,”Handwritten Character Recognition Using Gradient Features”, International Journal of Advanced Research in Computer Science and Software Engineering, Volume 2, Issue 5, May 2012. • Vinita Dutt, Sunil Dutt, “Handwritten Character Recognition Using Artificial Neural Network”, Advances in Computing: 2011; 1(1): 18-23. • Rahul Kala, HarshVazirani, Anupam Shukla, RituTiwari, “Offline Handwriting Recognition”, International Journal of Computer Science issues, volume 7, March-2010. • Dinesh Dileep, “A Feature ExtractionTechnique Based on Character Geometry for Character Recognition”. • Alexander J. Faaborg, “Using Neural Networks to Create an Adaptive Character Recognition System”, Cornell University, Ithaca NY, (May 14, 2002) • Swapnil A.Vaidya, Balaji R. Bombade “A Novel Approach of Handwritten Character Recognition using Positional Feature Extraction”, IJCSMC,Vol. 2, Issue. 6, June 2013. • Sheng Wang “A Review of Gradient-Based and Edge-Based Feature Extraction Methods for Object Detection”, Computer and Information Technology (CIT), 2011 IEEE 11th International Conference.
- 24. Thank You By: IndujaV (1PI12IS035) Medha.U.Kumar (1PI12IS055) Meghana Kantharaj (1PI12IS056)