Nigerian Language Dataset (Wa-Zo-Bia)

This is a dataset that contains the alphabets of the most common Nigerian languages from start to finish and can be used for character recognition. It was recorded physically and has been binarized, while some has not. The handwriting of 50 students was captured for both uppercase and lowercase for each of the languages. The dataset file: This file contains the raw images of the dataset; that is why it is the largest file. The binary file: This contains the raw data converted into binary format with a threshold of 210. This is why it is the smallest file. The sorted file: This file contains the sorted images, i.e., a folder was created for all the 'A' alphabets and so on till 'Z'. That is why it is different from the binary file. All you have to do is download the one you choose to use, and then unzip. The resized file: This contains all the images that have been resized to a specific dimension. Due to the existence of different contributors to the datasets, there is a variation of files and images. Have fun making use of it ;-)