Vehicle Detection using Aerial Imagery using Deep Learning Algorithm

The project aims at analyzing the various aerial images and solving any particular field (Vehicle detection) or sector related problems

Objectives:

1. Using standard MUNICH and OVERHEAD dataset and choosing the appropriate one.
2. Classification of vehicles into various classes.
3. Extraction of features for relevant objects.
4. Use of methods like HOG, SURF etc.
5. Evaluating and proving the performance efficiency and accuracy.

Methodology

Results

Further Work - Built SIMULO (free hand Web based tool to work with images)

We came to idea of building an Simulation tool which can be used to build NEURAL NETWORK and analyze performance by manipulating parameters.