Language-Independent ORB (Oriented Fast & Rotated Brief) Algorithm for Handwritten Documents
Journal Title: International Journal for Research in Applied Science and Engineering Technology (IJRASET) - Year 2017, Vol 5, Issue 5
Abstract
Handwritten character recognition is a demanding task in the image processing because handwriting varies from person to person. And also handwriting styles, sizes and its orientation make it complex. Applications like, handwritten text in reading bank cheques, Zip Code recognition and for removing the problem of handling documents manually, digital data is necessary. Recognition of handwritten characters using either a scanned document, or direct acquisition of image using Mat lab, followed by the implementation of various other Mat lab toolboxes like Image Processing to process the scanned or acquired image. Here OCR block diagram explained that how character are recognize accurately. Many feature-based algorithms are well-suited for character recognition like like SIFT, Language Independent Text-Line Extraction, Thresholding, Robust, Training, Ullman Algorithm, Structured Learning, ORB(oriented fast & rotated brief), SURF. But Oriented FAST and Rotated BRIEF (ORB) is a very fast binary descriptor which is faster than Scale-invariant feature transform (SIFT), it can be verified through experiments. Fast key point detector and BRIEF descriptor are important because of they have best performance and resonable cost. The recognize method for object recognition is Scale invariant feature transform (SIFT), which is very useful for feature extraction but it is computationally difficult due to its weighty workload required in local feature extraction and matching operation. Therefore for better performance and low complexity, ORB provides better solution. Recently there is a growing trend among worldwide researchers to recognize handwritten characters of many languages and scripts. Much of research work is done in English, Chinese and Japanese like languages .However, on Indian scripts the research work is lagging; most of research work is available is mainly on Devanagri and Bangala scripts. The work on other Indian scripts is in beginning stage. Therefore we have proposed offline recognition of handwritten characters of differen languagest characters.
Authors and Affiliations
Shweta Shevgekar, Mrs. Prof. M. S. Deole
Keywords
Related Articles
- EP ID EP24195
- DOI -
- Views 274
- Downloads 8
A Privacy-Preserving Personlized AdDissemination Based On Interest Aggregation and Piggybacking
Advertising on mobile devices has large potential due to the very personal and intimate nature of the devices and high targeting possibilities. we introduce a piggybacking for reduce the communication cost in privacy pr...
Survey on JPEG Image Compression Using DCT
This paper covers the analysis process of JPEG (joint picture expert group) standard which is based on the technique called discrete cosine transform (DCT). DCT is method that converts a graphic image as jpeg from spati...
Detailed Investigations on a A Hybrid Cloud Approach for Secure Authorized Deduplication
Data deduplication is one of important data compression techniques which is for eliminating duplicate copies of repeating data, and has been widely used in cloud storage in order to reduce the amount of storage space an...
Medical Algorithm for Bilirubin Identification with its Treatment Modes
The baby liver has a limited ability to process unconjugated bilirubin. Therefore, infant youngsters are slanted to a gathering of unconjugated bilirubin, and can make jaundice. Starting late, light transmitting diodes...
Object Detection Using Diversified Search by Measuring Objectness of Image Window
Object Detection and recognition is an important job in remote sensing and automation. Conventional object detection approaches are usually not sufficiently sturdy in addressing the variations of targets and typically s...