engleski

Diagram Reverse Engineering Using Computer Vision

This thesis describes the use of computer vision to recognize shapes and text from images depicting diagrams. The theoretical part of the thesis introduces basic concepts that serve as a foundation to the implementation of the practical part of the thesis. The practical part of the thesis consists of developing a Web application, which comprises a Web client and server. The client is a diagram editor capable of sending an image to the server to be interpreted. Images sent to the server are analyzed with OpenCV, a computer vision library, and a convolutional neural network model.

computer vision ; machine learning ; artificial intelligence ; REST API ; convolutional neural networks, OpenCV

nije evidentirano