Detection of Hindi Textual Characters from an Image
Pooja Patil, Kajol Patil, Anagha Kulkarni, Sahasra Iyer, Ira Natu · 2018
In the multimedia technology domain, Text Detection from an image is emerging as a fastest growing research and technology areas. The textual data in the images contain useful information for habitual explanation. The input to the system will be an image, particularly a colored one. The image will be processed to detect the area containing the text, the crucial features that uniquely identify the text characters will be detected and extracted and finally, the text is extracted into a text file. The text detection process begins with denoising of the image followed by converting it to a grayscale image followed by a binarization of that grayscaled image. Once grayscaled, the extracted text is written into a text file. The proposed model is robust to different font sizes, font colors, background colors. The performance of the presented system is demonstrated by showing the results for a set of images taken from various sources across the web.