Character Segmentation Research Articles

Automatic recognition of Indian vehicles license plates using machine learning approaches

A proposed method for automation of Indian vehicles number plate conversion is presented in this paper. Many vehicles are plying on Indian roads with varying in size and structure and hence there are a number of variations in number plates too. Although there are strict norms to be followed for number plate designs, however, it can be seen many deviations in actual number plates of many vehicles. It can be also observed that the number plates are written in many Indian languages, but the prescribed norm is Hindu-Arabic numerals with Latin letters which is prescribed by the licensing plate issuing authority. The number plate designs are varied according to types of vehicles, like, personal, commercial, public transport, etc. In order to design an automatic vehicle number plate, considerations of these factors are taken into account. The steps in the design include segmentation and character extraction from number plates and recognition of segmented characters in order to extract the complete information from number plates. The work in this paper is focused on Hindu-Arabic numerals with Latin letters. For feature extraction, Prewitt filter technique is used to extraction of characters from the vehicle number plates and connected component analysis is used for segmenting characters. Three different classifiers, namely, k-Nearest Neighbor, Artificial Neural Network and Decision Tree, are used in the experiments. Recognition accuracy up to 98.10% is achieved and is a promising factor for future research in this direction.

Two streams deep neural network for handwriting word recognition

Handwritten word recognition is one of the hot topics in automatic handwritten text recognition that received a lot of attention in recent years. Unlike character recognition, word recognition deals with considerable variations in word shape and written style. This paper proposes a novel deep model for language-independent handwritten word recognition. The proposed deep structure has two parallel stages for jointly learning character and word-level information. In the character-level stage, a weakly character segmentation method is performed and then applies a series of Long short-term memory (LSTM) layers for character-level representation. The word-level stage employs a series of convolutional layers for the shape and structure representation of the word. These representations are then concatenated and followed by a series of fully connected layers for jointly learning the words and the character-level information. Since the character segmentation is language independent and error-prone, the proposed deep structure only applies weakly separation scheme and does not rely on any character segmentation algorithm. Thus, it effectively utilizes character level representation without bounding on any language model. In the proposed methodology, we use two new data augmentation strategies based on a psychological assumption to increase the model generalization performance. Experimental results on five public datasets including Arabic, English and German languages demonstrate that the proposed deep model has a superior performance to the state-of-the-art methods.

Classification of Ancient Handwritten Tamil Characters on Palm Leaf Inscription Using Modified Adaptive Backpropagation Neural Network with GLCM Features

The core aspiration of this proposed work is to classify Tamil characters inscribed in the palm leaf manuscript using an Artificial Neural Network. Tamil palm leaf manuscript characters in the form of images were processed and segmented using contour-based convex hull bounding box segmentation. The segmented characters were transformed into two forms: Binary Coded Value and the Gray-Level Co-occurrence Matrix (GLCM) feature. The features extracted from the segmented characters were trained by the proposed method of the Modified Adaptive Backpropagation Network (MABPN) algorithm with Shannon activation function. Weight initialization plays an important role in the Backpropagation Neural Network, and hence Nguyen-Widrow weight initialization was introduced to initialize the weights instead of random weight initialization in the proposed method. The models evaluated are MABPN with Shannon activation function using Nguyen-Widrow weight initialization in two forms of input: Binary Coded Value and GLCM feature extracted values. The proposed method with GLCM features as input gave a promising result over binary coded transform.

An Automatic Traffic Rules Violation Detection and Number Plate Recognition System for Bangladesh

The traffic controlling system in Bangladesh has not been updated enough with respect to fast improving technology. As a result, traffic rules violation detection and identification of the vehicle has become more difficult as the number of vehicles is increasing day by day. Moreover, controlling traffic is still manual. To solve this problem, the traffic controlling system can be digitalized by a system that consists of two major parts which are traffic rules violation detection and number plate recognition. In this research, these processes are done automatically which is based on machine learning, deep learning, and computer vision technology. Before starting this process, an object on the road is identified through the YOLOv3 algorithm. By using the OpenCV algorithm, traffic rules violation is detected and the vehicle that violated these rules is identified. To recognize the number plate of the vehicle, image acquisition, edge detection, segmentation of characters is done sequentially by using Convolution Neural Network (CNN) in MATLAB background. Among the traffic rules, the following traffic signal is implemented in this research.

New species of caddisflies (Trichoptera: Psychomyiidae, Hydropsychidae, Leptoceridae) from Mekong tributaries, Cambodia.

Males of four new species of caddisflies from Kampong Speu and Siem Reap provinces, Cambodia [Psychomyia kimcheangi n. sp. (Psychomyiidae), Macrostemum siemreapensis n. sp. (Hydropsychidae), Leptocerus angkorwatensis n. sp. (Leptoceridae), and Trichosetodes kampongspeuensis n. sp. (Leptoceridae)] are described and illustrated. Psychomyia kimcheangi n. sp. can be distinguished from other species by characters of the inferior appendages. The main body of each inferior appendage has two branches that are separated near its base but are parallel and close together until the end. In Macrostemum siemreapensis n. sp. the dorsal tip of the phallus has no curved hooklet but instead a knot-like protruding structure in ventral view and looped-like structure in dorsal view. Leptocerus angkorwatensis n. sp. can be distinguished from others by the characters of long needled-like preanal appendages. Trichosetodes kampongspeuensis is distinguished from others by the characters of segment IX in ventral view, which is asymmetrical, and inferior appendages that are not divided and asymmetrical.

Automatic Number Plate Recognition System for the Detection of Unauthorized Vehicles During Government Imposed Lockdowns

Automatic Number Plate Recognition is a precise, error-free technology-integrated system used for the detection and recognition of number plates without human intervention of any sort. Considering the government-imposed lockdown in most of the countries of the world due to the rampant spread of the novel COVID-19 virus, we have attempted to present a fail proof method for the identification of unauthorized vehicles violating the government mandated norms in India without the intervention of police personnel. Furthermore, after the removal of the lockdown this system can be easily extended to detect transgression of traffic laws, automated commercial parking systems etc. The proposed morphological algorithm uses the application of Canny Edge Detector and Connected Component Analysis for the extraction of the plate followed by the usage of Bounding Box morphological technique for the segmentation of the individual characters within the plate. Recognition of the characters is done by Template Matching and Optical Character Recognition System.

Developing Arabic License Plate Recognition System Using Artificial Neural Network and Canny Edge Detection

In recent years, there has been expanding development in the vehicular part and the number of vehicles moving on the roads in all the sections of the country. Arabic vehicle number plate identification based on image processing is a dynamic area of this work; this technique is used for security purposes such as tracking of stolen cars and access control to restricted areas. The License Plate Recognition System (LPRS) exploits a digital camera to capture vehicle plate numbers is used as input to the proposed recognition system. Basically, the proposed system consists of three phases, vehicle license plate localization, character segmentation, and character recognition, the License Plate (LP) detection is presented using canny edge detection algorithm, Connect Component Analysis (CCA) have been exploited for segmenting characters. Finally, a Multi-Layer Perceptron Artificial Neural Network (MLPANN) model is utilized to identify and detect the vehicle license plate characters, and hence the results are displayed as a text on GUI. The proposed system successfully detects LP and recognizes multi-style Arabic characters with rates of 96% and 97.872% respectively under different conditions.

Online Persian/Arabic Writer Identification using Gated Recurrent Unit Neural Networks

Conventional methods in writer identification mostly rely on hand-crafted features to represent the characteristics of different handwritten scripts. In this paper, we propose an end-to-end model for online text-independent writer identification on Persian/Arabic online handwritten scripts by using Gated Recurrent Unit (GRU) neural networks. The method does not require any specific knowledge for handwriting data analysis. Because of the exclusive ability of deep neural networks, we just represented our data by Random Strokes (RS) representations, which are differential horizontal and vertical coordinates extracted from different handwritings with a predefined length. This representation is a context independent representation. Therefore, this writer identification at RS level is more general than character level or word level in identification systems, which require character or word segmentation. The RS representation is then fed to a GRU neural network to represent the sequence for final classification. All RS features of a writer are then classified independently, and in the final stage, the posterior probabilities are averaged to make the final decision. Experiments on KHATT database, which consists of online handwritings of Arabic writers, gave us 100% accuracy on 10 writers and 76% accuracy on 50 writers, which is much better than previous works on online Persian/Arabic writer identification.

Arbitrary oriented multilingual text detection and segmentation using level set and Gaussian mixture model

A pioneer conceptual combination of level set method and Gaussian Mixture Model (GMM) is presented in the described multilingual, arbitrary-oriented character segmentation. The method is a serial accomplishment processes of Gaussian low pass filter, single level 2 Dimensional Discrete Wavelet Transform (2D DWT) for better feature extraction and implemented level set method, k-means clustering algorithm with GMM to achieve a veracious character segmentation results by detecting great measure of true text region of an image. The proposed method segments a character chiefly by distinguishing the touching character constituents and also deals discontinuities presence in a character by using Laplacian of Gaussian filter and morphological bridge function in a intellectual way. The exhibited method was explored on Multi-script Robust Reading Competition dataset and on our privately collected graphical and handwritten multilingual, arbitrarily-oriented text images. The suggested method is compared with the well known multilingual and arbitrarily-oriented charter segmentation methods. The described method attains better segmentation outcomes when compared to the familiar functioning methods. Hence, the suggested method is highly suitable to consider as an improved, standard and procedural technique.

OHKWR: Offline Handwritten Kannada Words Recognition using SVM Classifier with CNN

In field of handwriting recognition, Robust algorithms for recognition and character segmentation are presented for multilingual Indian archive images of Devanagari and Latin scripts. These report basically suffer from their format organizations, low print and local skews quality and contain intermixed messages (machine-printed and manually written). In order to overcome these drawbacks, a character segmentation algorithm is proposed for kannada handwriting recognition. In this work, in initial steps we are obtained the segmentation paths by using the characters of structural property and also the graph distance theory whereas overlapped and connected character are separated. Finally, we are calculated results by using the SVM classifier. In proposed recognition of character, they are three new geometrical shapes based on new features such as center pixel of character is obtained by first and second feature and third feature is calculation purpose we are used in neighborhood information of text pixels. Benchmarking results represent that proposed algorithms have best work identified with other contemporary methodologies, where best recognition rates and segmentation are obtained.

Kannada Text to Speech Conversion System

The following paper describes the design of a system which does text to speech generation for one of the regional language’s Kannada. The printed document of Kannada text is given as input to the system, the system then converts the document to an image format. Pre-processing is done to stabilize the intensity of the images and clear the artifacts. This process boosts the precision and interpretability of an image. Optical Character Recognition (OCR) is used to unsheathe the segmented characters from a particular image and are matched with the characters that have been stored in the dataset. Once the matched characters are extracted it is stored in a suitable format and then the TTS engine is deployed to convert the saved Kannada characters to a speech format. The obtained speech output corresponds to the characters which are collected after processing the input text.

An Effective Approach for Noise Robust and Rotation Invariant Handwritten Character Recognition Using Zernike Moments Features and Optimal Similarity Measure

ABSTRACT Zernike moments (ZMs) are very effective orthogonal rotation invariant moments. Conventionally, the magnitudes of ZMs are used as feature descriptors and the Euclidean distance is used as a classifier. Recently, a few classifiers based on ZM magnitude and phase have been developed which are reported to be very effective in pattern matching problems. One such a recently developed similarity measure, known as optimal similarity measure, is known to provide very good performance over the ZM magnitude-based Euclidean distance measure in pattern recognition problems, especially under noisy conditions. In this paper, we investigate the conventional magnitude-based similarity measure and the new similarity measures including the optimal similarity measure and compare their performance on segmented handwritten characters and numerals. It is observed that the performance of optimal similarity measure is far better than all other similarity measures. Its performance is very much better than other similarity measures even under very high noisy condition. However, it is slow owing to the optimization of the process involved in its computation. Therefore, we also propose a fast algorithm for its computation and reduce its time complexity. Detailed experimental results are provided to support the above observations.

Random-Positioned License Plate Recognition Using Hybrid Broad Learning System and Convolutional Networks

This paper proposes a framework combing a fully convolutional network with broad learning system for license plate recognition. The fully convolutional network, which is designed as a pixel-level two-class classification method, is proposed for random-positioned object detection by the fusion of multi-scale and hierarchical features. For character segmentation, a trained AdaBoost cascade classifier is employed to locate a key character representing for an administrative area. We design a symmetric region horizontal projection method to estimate the license plate slant angles, and an approach based on vertical projection without hyphens to solve the problem of touching characters. For character recognition, the broad learning system with stacked auto-encoder of mapped feature nodes is proposed, and two structures are explored to recognize letters and digits, respectively. Experiments conducted on Macau license plates show that the proposed method outperforms some state-of-the-art approaches. The compatibility and generality can be expected by applying the proposed method to other regions or countries.

Image pattern recognition in identification of financial bills risk management

The automatic identification system of financial bills needs to have high recognition rate, high anti-interference and real-time to ensure its recognition effect. Based on the image recognition theory, this study uses the differential projection method to define the boundary of the bill. The horizontal projection gradation of the boundary area of the bill character line is significantly reduced from large to small, and the horizontal difference projection of the grayscale image can be performed to locate the boundary of the bill image. The word height of bills studied in this paper is constant, the character line spacing is equal, and the horizontal differential projection is used to realize row positioning. The main process of the system is to first select the check image, and then the software part realizes the process of image analysis, preprocessing, character segmentation, feature extraction, character recognition, etc. Finally, the recognized amount is output to the interface. It is proved by experiments that the recognition rate of this algorithm is high, which can provide theoretical reference for subsequent related research.

Comparative Analysis of Segmentation and Recognition Techniques for Offline Handwritten Words

A Pre-processing is the initial and vital phase in optical character recognition is the Pre-processing. Segmentation deals with the extraction of individual component from a document image. Number of techniques like projection profile, connected components, gaps between characters/components is reported in the literature for component extraction followed by feature extraction and recognition of the individual component. These techniques gives good results if components are isolated but fails if components are touched, shadowed or skewed. A novel technique is required to address such issues to enhance the recognition rate. The problem of segmentation for Roman script cursive handwriting is addressed by various authors but not enough addressed for Indian script especially Devanagari script. This paper is a review which is confined to offline handwritten script domain. It attempt to review various techniques for character segmentation considering touching characters for offline handwritten words in Devanagari script and scripts sharing similar characteristics (like Bangla, Gurumukhi), database used and their accuracy reported in the literature.

Word and Chracter Segmentation in Devnagari and Odia Script – A Comparitive Analysis

Optical Character Recognition has been an active research area in computer science for several years. Several research works undertaken on various languages in India. In this paper an attempt has been made to find out the percentage of accuracy in word and character segmentation of Hindi (National language of India) and Odia is one of the Regional Language mostly spoken in Odisha and a few Eastern India states. A comparative article has been published under this article. 10 sets of each printed Odia and Devanagari scripts with different word limits were used in this study. The documents were scanned at 300dpi before adopting pre-processing and segmentation procedure. The result shows that the percentage of accuracy both in word and character segmentation is higher in Odia language as compared to Hindi language. One of the reasons is the use of headers line in Hindi which makes the segmentation process cumbersome. Thus, it can be concluded that the accuracy level can vary from one language to the other and from word segmentation to that of the character segmentation.

SNRNet: A Deep Learning-Based Network for Banknote Serial Number Recognition

The banknote serial number recognition (SNR) plays an important role in the banking business and attracts much attention recently. However, most of the existing SNR methods take character segmentation and character classification as two separate steps, so that the accuracy of SNR heavily relies on the character segmentation, which is a challenging problem due to complicated background and uneven illumination. In this paper, the SNR is cast into a sequence prediction problem, which integrates such two steps into a unified network, and we propose a deep learning-based serial number recognition network, which can be trained end-to-end to avoid the preliminary character-segmentation with three steps as follow. First, the improved convolutional neural networks are employed to extract the feature sequence of the input image. Second, the feature sequence is used as an input to the bidirectional recurrent neural networks (BRNNs), where the character segmentation is not required. Finally, the label recognition is implemented using the connectionist temporal classification to decode the BRNNs’ output. The experimental results demonstrate that the proposed method outperforms the state-of-the-art methods in both accuracy and efficiency: it achieves character and serial number recognition of the renminbi (RMB) with accuracies 99.96% and 99.56%, respectively.

Efficient framework with sequential classification for graphic vehicle identification number recognition

In vehicle monitoring, recognizing graphic vehicle identification number (VIN) on the car frame is a particularly important step. While text recognition methods have made great progress, automatic graphic vehicle VIN recognition is still challenging. In VIN images, the VIN text is engraved on the car frame, with complex background and arbitrary orientation, which make it extremely difficult for recognition. We propose an efficient framework for recognizing rotational VIN. First, combining lightweight convolutional neural network and per-pixel segmentation in the output layer, we achieve fast and excellent VIN detection. Second, we take the VIN recognition task as a sequential position-dependent classification problem. By attaching sequential classifiers, we predict VIN text without character segmentation. Finally, we introduce a VIN dataset, which contains 2000 raw rotational VIN images and 90,000 horizontal VIN images for validating our framework. Experiments results show that the framework we proposed achieves good performance in VIN detection and recognition. By automatically identifying the VIN, we can quickly confirm the vehicle’s identity and help vehicle monitoring and tracking.

Design of license plate recognition system based on capsule network

With the development of automobile industry, the license plate, as the unique identifier of the vehicle, plays an important role in vehicle management. License plate recognition technology is mainly realized by image processing technology. It mainly includes image preprocessing, license plate positioning, character segmentation and other processes. In the actual process of license plate recognition, it is often impossible to accurately recognize the license plate due to the positional relationship such as angle. Therefore, this paper proposes a license plate recognition program based on the capsule network. First, through image preprocessing, license plate positioning, and character segmentation, and then use capsule neural network for training and simulation in order to achieve the object of accurate character recognition.

Development of Iraqi License Plate Recognition System based on Canny Edge Detection Method

In recent years, there has been expanding development in the vehicular part and the number of vehicles moving on the road in all the sections of the country. Vehicle number plate identification based on image processing is a dynamic area of this work; this technique is used for security purposes such as tracking of stolen cars and access control to restricted areas. The License Plate Recognition System (LPRS) exploits a digital camera to capture vehicle plate numbers is used as input to the proposed recognition system. Basically, the developing system is consist of three phases, vehicle license plate localization, character segmentation, and character recognition, the License Plate (LP) detection is presented using canny Edge detection algorithm, Connect Component Analysis (CCA) have been exploited for segmenting characters. Finally, a Multi-Layer Perceptron Artificial Neural Network (MLPANN) model is utilized to recognize and detect the vehicle license plate characters, and hence the results are displayed as a text on GUI. The proposed system successfully identified and recognized multi_style Iraqi license plates using different image situations and it was evaluated based on different metrics performance, achieving an overall system performance of 91.99%. This results shows the effectiveness of the proposed method compared with other existing methods, whose average recognition rate is 86% and the average processing time of one image is 0.242s which proves the practicality of the proposed method.