Segmentation of Multiple Touching Hand Written Devnagari Compound Characters

Yawalkar, Prashant Madhukar; Kharat, M. U.; Gumaste, Shyamrao V.

doi:10.4018/978-1-5225-5775-3.ch008

Cited by 4 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…L is a blurred image, G is the Gaussian Blur operator, I is the SAR image, a,b are the location coordinates, and z is the scale parameter. Difference of Gaussian (DoG) images are roughly equal to the Laplacian of Gaussian (LoG) images as shown in Figure 5 (Yawalkar et al 2018). Thirdly, we discover the maxima and minima in the DoG images created in the previous step.…”

Section: Registration With Siftmentioning

confidence: 89%

Registration-based change detection for SAR images

Gharbia

Amin

Mousa

et al. 2020

NRIAG Journal of Astronomy and Geophysics

View full text Add to dashboard Cite

This paper presents an efficient change detection approach for Synthetic Aperture Radar (SAR) images. The basic idea of this approach is to use the log ratio of the two images for change detection after being registered with Scale-Invariant Feature Transform (SIFT). These two images are a reference image and another image for the same area acquired at a different time. The log ratio variations include changes in certain areas corresponding to the natural changes in the test image. Usually, SAR images contain some sort of noise. So, there is a need for a denoising process prior to estimating the log ratio to enhance the change detection results. A segmentation process is performed on the test image based on the log ratio values. Large values in the log ratio image correspond to detected changes in the test image. Simulation results on SAR images for a region of Jeddah demonstrate the success of the proposed approach.

show abstract

Section: Registration With Siftmentioning

confidence: 89%

Registration-based change detection for SAR images

Gharbia

Amin

Mousa

et al. 2020

NRIAG Journal of Astronomy and Geophysics

View full text Add to dashboard Cite

show abstract

“…Preprocessing focuses on deduction of the noise resulting due to the scanner quality used for capturing the image. In order to make the further recognition process smooth apart from noise reduction, other processing activities like thinning, normalization and segmentation [19], [27] of image are also carried out. Normalization involves resizing of characters for stroke width, slant, slope, height of the characters.…”

Section: Basic Stages In Character Recognitionmentioning

confidence: 99%

Effective Thinning Algorithm for Recognition of Hand Written Devnagri Compound Characters Using Neural Network

Yawalkar¹,

Kharat²

2018

International Journal of Applied Engineering Research

View full text Add to dashboard Cite

Handwriting always works as a primary tool in the world of communication. Various techniques have been and are developed in order to document the handwritten text. There are certain exceptional techniques that describe the environment of handwritten scripts and further convert it into electronic data by implementing various algorithms. Devnagri is widely used for most popular and commonly used languages like Marathi and Hindi We have proposed a neural network based hand written Devnagri compound character recognition system using a novel rule-based thinning algorithm that improves the overall performance of the system. The distinctive feature of our thinning algorithm is that it is a rotation invariant method which thins characters to their central lines, helping in preserving shape of the character. The system applies different rules concurrently to each pixel in the image which results into symmetrical thinning and improvement in the overall speed of the system. Results obtained for the system prove that the system is efficient enough to preserve the topology of the compound characters written in Devnagari that further helps in improving the accuracy of the recognition system.

show abstract

“…Segmentation which is a decisive activity in the process of character recognition decomposes the input image into sub images by means of separating lines and words. Different techniques for decomposition of multiple touching handwritten Devnagri characters were presented [8]. Extracting features from the input image in the sense of information that must be similar for similar images and distinct for other images.…”

Section: Introductionmentioning

confidence: 99%

A Hybrid Approach for Recognition of Hand Written Devnagri Compound Characters

Yawalkar¹,

Kharat²

2019

AJCST

View full text Add to dashboard Cite

Being an effective tool in the world of communication, numerous techniques have been developed for documenting the handwritten text. Few of the exceptional techniques describe the environment of handwritten scripts and further convert it into electronic data by implementing various algorithms. Devnagri is one of the widely used scripts for most popular and commonly used languages like Marathi and Hindi. Recent development in the field of handwritten character recognition based on different methodologies like neural network, fuzzy logic, and deep neural networks has shown remarkable improvement in character recognition accuracy from 75% to 96%. We propose a fuzzy-Neural hybrid approach for recognition of hand written Devnagri compound character that uses a rotation invariant rule-based thinning algorithm as one of the major pre-processing activity. Thinning the characters to their central line, preserving the shape of the character are the distinctive features of thinning algorithm. Concurrent application of different rules to each pixel of the character image results into symmetrical thinning as well as improves the overall speed of the system. The system is trained using Neural Network where the weights are optimized using fuzzy rules improving the accuracy of the system.Results obtained for the fuzzy-neural based system with thinning helps in preserving the topology of the characters written in Devnagri and prove that accuracy of the system has stabilized in the band of 92-97% which was fluctuating in the band of 89-94% for the previously implemented systems. The system also shows a substantial improvement in accuracy for recognition of compound characters in comparison with our previously implemented system.

show abstract

Segmentation of Multiple Touching Hand Written Devnagari Compound Characters

Cited by 4 publications

References 5 publications

Registration-based change detection for SAR images

Registration-based change detection for SAR images

Effective Thinning Algorithm for Recognition of Hand Written Devnagri Compound Characters Using Neural Network

A Hybrid Approach for Recognition of Hand Written Devnagri Compound Characters

Contact Info

Product

Resources

About