DESCRIPTION:

Text is ubiquitous in man-made environments, and most daily activities involve reading and understanding written information around us. Our current technology permits an end-to-end recognition of structured text in images such as numeric values, dates, amounts, serial numbers, etc. By means of the latest advances in deep learning, we have developed a technology for single-shot text interpretation in images, without any text segmentation step, and ignoring irrelevant textual content resulting to a fully automatic, high precision, smart and specialised reading system.