Inside Images: Grey-Scale by ACM
Participants of the Association for Computing Machinery’s Grey-Scale learnt to look at pictures on a whole new level as the presenters walked them through the basics of OpenCV. OpenCV is an open source computer vision and machine learning software library that can be used for everything from interactive art to advanced robotics.
The first day of the workshop started with an explanation of what pixels are and how they’re stored, accessed, and modified, in addition, the concepts of edge detection, colour identification, and shape recognition were also touched upon. On the second day, the presenters demonstrated how the software can differentiate between different objects even when they have the same colour and shape. This is done by using a Linear Binary Pattern that generates a grayscale image by comparing pixels with all their neighbouring ones to get the texture of the object. Features like Optical Flow, that achieves motion detection by comparing the positions of pixels with respect to each other in short intervals of time, were also explored. Insights into facial recognition and its workings were also provided, with explanations on how characteristics like a person’s eyes could be distinguished using the contrast between the black iris and the surrounding white.
The fact that only half a dozen participants turned up on the second day did not disappoint the organisers because, as ACM President Rakshit Thakkar explained, image processing was a concept usually studied only at the post-graduate level. This also meant that those who did sit through the session were those who were genuinely interested in the topic and keen on learning. The workshop engaged the participants and involved them in the discussion throughout, keeping the whole event lively and interactive. The participants walked in without needing any prerequisites, and finished the session with enough skills for simple image processing, like identifying small objects using OpenCV.
The workshop kept the participants engaged and involved them in the discussions, ensuring that it did not turn into a monotonous lecture. Over the course of the weekend, the participants, who had walked in even without any prior experience in the area, acquired enough skills for simple image processing, such as identifying small objects.