Abstract:
The project Automated Recognition of Components from Hand Drawn Circuit Diagrams
developed to recognize circuit components from hand drawn circuit diagrams on paper.
The hand drawn circuit diagram will be acquire as a scanned image to the application.
Each component will be selected by user from circuit diagram to recognize. From which
Histogram of Oriented Gradient (HOG descriptor) will extract features to match. A data
set of trained features will also be maintained to match features from. Then Euclidean
Distance will be calculated from trained features to get nearest neighbors. This way most
repeating of nearest 5 components will be selected. Which will be labeled as a result in
circuit diagram. User will also have a short description of recognized components.
