Hyeonho Shin

In order to solve the drone’s flight time problem, an attempt is made to fuse with an airship for better flight time. The research was conducted with a grant of $1,300 from the Hanyang University Smart Teaching and Learning Center.

To address the police’s lack of search personnel, we have developed a drone to search for lost children given their image data. At this time, a drone equipped with Jetson TX2 was developed, and the widely used YOLO Network was run on the drone to separate people and objects in real time.

Leveling process is essential in modern 3D printers, but there is a problem that the sensor needs to be calibrated frequently because the nozzle and sensor from which filament is extruded are separated. To solve this problem, we invented a method of integrating a pressure measuring sensor into the bed to position the correct position of the nozzle.

In the conventional lane detection algorithm, it was difficult to extract lane features due to saturated pixels under strong sunlight. This was a major cause of accidents in Tesla’s autonomous vehicles. In this study, we proposed tanukiNet, a lane detection neural network that is resistant to external environmental changes, and verified it by mounting it on Jetbot.

To make it easier for newcomers to get the information they need about Hanyang University, I built, registered and distributed an unofficial wiki.

Using the Raspberry pi 3 and Octoprint library, 3D printing at home is possible without going in a club room, enabling remote control and webcam monitoring.