Handwritten optical character recognition (OCR) is a noteworthy research region because of its sensitivity in segmenting the character which increments on account of MARATHI script because of modifiers and compound characters. This paper gives a streamlined OCR framework for handwritten MARATHI text document classification and recognition system. To deal with a vast measure of features, the support vector machine (SVM) assumes a noteworthy part which was likewise used for the classification reason. In this paper, we display a projection profile segmentation technique which generates less error. The Curvelet Transform (CT) to be exceptionally efficient and hearty to get the feature characters from the pre-processed image. The extracted feature sets are decreased by Principle Component Analysis (PCA) algorithm. After the feature extraction process, the Adaptive Cuckoo Search (ACS) algorithm is used for the optimization procedure. Here, the written by hand MARATHI script was segmented flexibly in three levels; (1) line segmentation, (2) word segmentation and (3) character segmentation. The preprocessing was finished utilizing different morphological operations. The experimental results show that, the performance of the proposed technique is assessed in view of the accuracy, sensitivity, precision, recall and F-score. Compared with the existing Fire Fly Selection (FFS) and Bat Selection (BS) approach, the proposed method has 99.36% accuracy, 90% sensitivity, 91% precision, 89.51% recall, 99.67% specificity and 89.93% F-score. The proposed approach is actualized using MATLAB and the realtime Marathi character datasets are used for our examination.

No abstract

No abstract

ABSTRACT:In OCR domain, it is now widely accepted that a single feature extraction method and single classification algorithm can't yields better performance rate. It is therefore, a compound feature extraction approach based on structural analysis for recognition of offline handwritten Marathi vowels is proposed. Though, Moment invariant technique is well experimented by many researchers, an attempt is made to enhance the existing results by extracting various supportive features like affine invariant moments, image thinning, structuring the image in box format, etc. These features are independent of size, slant, orientation, translation and other variations in handwritten characters. 5 samples of each vowel from 25 different people have been sampled and database was prepared. After segmentation, an individual image is resized to 50X50. 33 different features were evaluated for each character. The Fuzzy Gaussian Membership Function has been adopted for classification. The main objective of the paper is to test the possibility of using the MI, AMI combination of both for recognition of Handwritten Marathi vowels. The results show the satisfactory performance rate.

Now-a-days there are many new methodologies required for the increasing needs in newly emerging areas, with these methodologies there are many techniques are present for the character recognition of handprint Devnagari, Bangla, Tamil, China etc. Also there is lot of work is done for the printed material but it is only limited for laboratory. But it has not been used practically. So in this paper, proposed a Minimum distance classifier technique for OCR System of printed as well as scanned newsprint Marathi script.