Computer Vision News - August 2021

DIDA and DIGITNET Amir Yavariabdi is an assistant professor in KTO Karatay University in Turkey. Hüseyin Kusetogullari is a Senior Lecturer in the Department of Computer Science at the Blekinge Institute of Technology in Sweden. They are speaking to us about DIDA, a new historical handwritten digit dataset, and DIGITNET, a deep handwritten digit detection and recognition method. DIDA is the largest image-based historical handwritten digit dataset , with 250,000 single digits and 200,000 multi-digits in RGB color space, collected from historical handwritten document images. Around 75,000 document images were used, written by different priests with different handwriting styles in Sweden in the 1800s and 1900s. DIGITNET uses a YOLO-based detection algorithm network , combined with an external recognition algorithm for higher accuracy, to detect and collect these digits inside the document images. Amir Yavariabdi Hüseyin Kusetogullari 28 Computer Vision Project