Saudi Currency Dataset|纸币识别数据集|计算机视觉数据集

Automatic detection and recognition of banknotes could be a highly important technology for people with visual impairments as well as banks, providing easy administration for dealing with diverse paper currencies. With this purpose in mind, we created the Saudi riyal banknotes dataset with annotations. We utilized a smartphone camera to create our new dataset consist of Saudi riyals. All market-acceptable currency notes are used, including 1, 5 riyal notes, 10 and 20 riyal notes, 50 riyal notes, and new 100, 200, and 500 riyal notes. We capture photos from multiple viewpoints to extract additional information from the image. For each Saudi riyal, we employ twelve distinct angles. Our dataset includes 2000 labeled photos. For annotating our photos in our dataset, we utilize Pascal Visual Object Classes. The dataset is made up of "train" and "test" photos in a 90:10 ratio.