Gurmukhi handwritten word conversion using feature extraction using deep learning

Digitalization has resulted in creation of on-line repository off all government documents, which is as-sembled in fraction of seconds. But in the various fields of handwritten manuscript such as for a postal documentation it's considered to bit difficult task to recognize the handwritten city names to reach its destination location. In this study, recognition system for Gurmukhi hand-written names is presented. The features of the Gurmukhi script are discussed prior to actually empha-sizing on the features of word recognition system works on Gurmukhi. The handmade, printable, and digital characters and words identification for Punjabi language are taken into account for clarity and although includes the conversion of normal handwritten data into grey scale images that can be easy for the system to recognize. After a full analysis of all relevant articles, a detailed as-sessment of the feature extraction phase is performed, along with a histogram of the results.