The wonderful world of image processing

Image processing is a very broad and deep area of research in the modern day and age. As cliche as it may sound, a picture speaks a thousand words. In computing terms, the pixels in an image can convey a TON of information. The data stored in each individual pixel, when strung together, can tell a wonderful story.

In a search and rescue scenario, what do we actually need? The landscape may be completely destroyed, and SAR efforts are carried out to look for survivors and reclaim bodies, if any. Having humans to carry this out is not as efficient as having a machine carry this out. By having a drone to fly through an area, map it out, and also search for survivors using a mere camera is very possible and at the same time very powerful.

For the scope of this project, I have employed the use of OpenCV to carry out all the necessary image processing. All the programming was done in Python 2.7. The reason for not making use of Python 3.0 or later versions, or even C++ is due to the fact that our control is being accomplished via the use of Dronekit, which has a stable Python 2.7 release. As such, to simplify inter-script communication and to save on any compatibility issues, we have had to consider very carefully the versions of the various APIs and libraries we use and also the drawbacks and benefits of each version.

Nonetheless, I do look forward to the adventure in computer vision!