Simultaneous Detection And Recognition of Named En

Polar Rain proposes to study, and investigate approaches to develop a new data and computational models combining image-based features with language model to extract name entities in degraded Arabic printed and handwritten documents. The proposed approach will use robust shape features and computational model that simultaneously clean and restore document images and detect named entities. The proposed data models are specifically designed taking named entities characteristics into consideration. In phase-I, we will investigate how to design such data model which contains both image and language features together for complex degraded document images. Our computational models will use generative and descriptive models to extract names from recognition results that may contain significant error or partial recognition. The proposed approach will work both in handwritten and printed Arabic scripts. We will design a flexible software architecture so that the solution can be extended to other similar languages, such as Pashto, without extensive work and allows 3rd party OCR to be integrated with the system. We will evaluate and conduct experiments to measure feasibility and superiority of proposed solution over exiting NEE solution.