Optimally configured convolutional neural network for Tamil Handwritten Character Recognition by improved lion optimization model

Abstract

In recent data, Optical character recognition (OCR) systems have laid hands in the field of most popular language recognitions. Unlike other languages, the Tamil language is more complex to recognize, and hence considerable efforts have been laid in literature. However, the models are not yet well-organized for precise recognition of Tamil characters. Thus, the current research work develops a novel Tamil Handwritten Character Recognition approach by following two major processes, viz. pre-processing and recognition. The pre-processing phase encloses RGB to grayscale conversion, binarization with thresholding, image complementation, morphological operations, and linearization. Subsequently, the pre-processed image after linearization is subjected to recognition via an optimally configured Convolutional Neural Network (CNN). More particularly, the fully connected layer and weights are fine-tuned by a new Self Adaptive Lion Algorithm (SALA) that is the conceptual improvement of the standard Lion Algorithm (LA). The performance of the proposed work is compared and proved over other state-of-the-art models with respect to certain performance measures.

uTHCD: A New Benchmarking for Tamil Handwritten OCR

“…The availability of a comprehensive database can also help to objectively assess the true efficacy/limitations of SOTA algorithms [18]. Developing a robust Tamil OCR system necessitates a large and extensive database [10]. As mentioned earlier, there is only a single standardized database developed by HP labs India, available for the Tamil OCR domain [18].…”

“…Tamil Handwritten character recognition is one such challenging research topic for close to 4 decades now [7], [8]. Despite this, it continues to offer many challenges which keeps the research community active even till date [9], [10]. The OCR of printed documents has been considered a problem solved par human accuracy or even better [11]- [13].…”

“…This database is publicly available and was created using online pen coordinates by applying a simple piece-wise interpolation process. Most of the work reported in Tamil handwriting recognition has relied mainly on this database [9], [10]. However, for more objective study and to evaluate the true efficacy of stateof-the-art(SOTA) algorithms, having multiple standardized publicly available datasets exhibiting a variety of inherent writing variations can only add more value to the resulting experimental analysis.…”

Enhancing Offline Tamil Handwritten Character Recognition Using Optimal Newton Algorithm Based Deep Convolution Extreme Learning Model

“…In order to gain better understanding, the highlights of few research works are delineated in the following section. Lincy et al (2020) demonstrated an optimal configuration of convolutional neural network for Tamil Handwritten Character Recognition using an improved lion optimization algorithm. The evaluation measures employed in recognizing the Tamil characters were Accuracy, sensitivity, specificity and precision.…”

“…This section discusses the comparative results of various simulation metrics namely accuracy, sensitivity, specificity and F1-score for various existing Tamil character recognition approaches namely Neural network base Elephant herding optimization (NN-EHO) (Kowsalya and Periasamy, 2019), Convolution Neural Network-based Improved Lion optimization (CNN-ILO) (Lincy et al 2020), Convolution neural networks (CNN) and Modified Convolution neural networks (MCNN). Fig.…”