Text Segmentation and Recognition for Enhanced Image Spam Detection

The book "Text Segmentation and Recognition for Enhanced Image Spam Detection" offers a comprehensive analysis of the challenges in email spam detection, particularly regarding text classification and categorization. The author, Mallikka Rajalingam, presents an efficient technique for spam detection based on a combination of character segmentation and recognition as well as classification. The proposed solutions include innovative approaches for extracting text characters from images, supported by a segmentation process that utilizes Discrete Wavelet Transformation (DWT) and slant detection. Additionally, methods for text character recognition and visual feature extraction are described, based on improved contour analysis. The classification of the extracted text features is carried out using optimized K-Nearest Neighbor (KNN) and Support Vector Machine (SVM) techniques. Finally, the performance of the proposed method is validated using various metrics such as sensitivity, specificity, precision, and accuracy.