Project Overview

Brahmilator

Proposed system has 5 different core modules.

1. Pre - Processing Module
2. Character Segmentating Module
3. Classification Module
4. Words Generating Module
5. Translating Module

Pre - Processing Module

Whenever consumer takes a picture from the mobile camera, first it goes through the Pre - Processing Module. This is where consumer gets to clear the image and ready for the segmentation module. Brahmilator provides 5 different operations to pre-process the image. Following OpenCV operations have been successfully implemented through a React-Native Bridge.

1. Threshold
2. Erosion
3. Dilation
4. Opening
5. Closing

Character Segmentation Module

Segment lines of Brahmi characters in a given image using horizontal projection then Separate each character of each line by cropping on the drawn contour around the character. This process functions in the back-end and does not interact with the front-end at all. Finally, all segmented images are sent to the classification model to be classified.

Classification Module

Words Generating Module

Translating Module