Segmentation-free Connectionist Temporal Classification loss based OCR Model for Text Captcha Classification

TL;DR

The paper tackles text-based CAPTCHA recognition under distortions and variable lengths. It introduces a segmentation-free OCR approach that combines CNN-based spatial feature extraction, RNN-based sequence modeling, and CTC loss to align predictions with ground truth without explicit character segmentation. On a public CAPTCHA dataset, the method achieves 99.80% character-level and 95% word-level accuracy, surpassing several state-of-the-art methods. This work demonstrates robust, scalable CAPTCHA recognition with practical implications for security systems and guides future research in data augmentation, transfer learning, and multimodal approaches.

Abstract

Captcha are widely used to secure systems from automatic responses by distinguishing computer responses from human responses. Text, audio, video, picture picture-based Optical Character Recognition (OCR) are used for creating captcha. Text-based OCR captcha are the most often used captcha which faces issues namely, complex and distorted contents. There are attempts to build captcha detection and classification-based systems using machine learning and neural networks, which need to be tuned for accuracy. The existing systems face challenges in the recognition of distorted characters, handling variable-length captcha and finding sequential dependencies in captcha. In this work, we propose a segmentation-free OCR model for text captcha classification based on the connectionist temporal classification loss technique. The proposed model is trained and tested on a publicly available captcha dataset. The proposed model gives 99.80\% character level accuracy, while 95\% word level accuracy. The accuracy of the proposed model is compared with the state-of-the-art models and proves to be effective. The variable length complex captcha can be thus processed with the segmentation-free connectionist temporal classification loss technique with dependencies which will be massively used in securing the software systems.

Figures (5)

• Figure 1: Proposed Model
• Figure 2: Sample data
• Figure 3: Data Pre-processing Steps
• Figure 4: Training of model
• Figure 5: Loss Curve