Tuesday


Tuesday kicked off bright and early with breakfast at the hostel before heading to the PXL campus. The day was packed from the start. It began with a full dive into computer vision.

The morning sessions were focused on understanding the basics of image processing and the tools we’d be using for the rest of the week. We were introduced to OpenCV, one of the most widely used libraries for computer vision. The presenters walked us through how images can be filtered, transformed, and analyzed to detect shapes, colors, and edges. It was surprising how much goes on behind the scenes of something as simple as detecting a line or recognising an object in a photo.

After lunch, we moved on to project ideation. This was where teams brainstormed how to apply computer vision in the context of the week’s challenge. Ideas started forming quickly, and it helped to share thoughts across groups to get some early feedback.

The rest of the afternoon was all about trying things out. We opened the Gazebo simulator, dropped in a Prius model, and started working with its front camera feed. Using OpenCV, our goal was to make the car respond to visual input. It wasn’t easy, some of us spent ages just trying to get clean edge detection or figuring out HSV thresholds. But slowly, things started clicking into place.

Some teams got their virtual cars moving along the racetrack, others were still debugging by the end. Either way, it felt like a solid step forward from theory to practice.

By 5pm, we wrapped up and headed back, a bit tired but definitely more confident than the day before.

Georgy Gurevich

Reacties

Populaire posts van deze blog

The First Day

Thursday