The problem Field Data and Automated Feature analysis solves

For understanding the relevance and extracting information from crowd sourcing photos for Geospatial technology, object detection and text recognition will be used which will enable the extraction of information.

Challenges I ran into

Integrating the neural networks with Interface for text extraction from images which are not in orthogonal plane with respect to our point of view or reference. Used ResNet to extract the text from these type of images by changing the orientation.