Video & Image Engineering Lab

(Prof. Whoi-Yul Kim)

The research field of Prof. Whoi-Yul Kim’s laboratory can be largely divided into Object/Pattern Recognition, Intelligent Surveillance, HCI (Human Computer Interface), and Intelligent Vehicle.

Object/Pattern Recognition is a technology that recognizes specific objects or objects with a specific pattern in an image. Some examples of object detection technology are face detection, eye detection, and pupil detection. Face recognition, defect recognition, character recognition and bone age estimation are some examples of object recognition technology.

Intelligent surveillance is a field that improves the existing system of simple camera footage recording to intelligently detect and track moving objects in an image. This field also involves detection of intrusions, loitering individuals, and unattended objects. There is a variety of application that benefits from the intelligent surveillance system. Among them are the ability to count the number of customers in a large shopping mall area, effectively plan and announce evacuation routes in the event of an emergency such as fire and setting store rent payment differently depending on the average number of customers in a region.

HCI is a field that develops an effective interface between the user and the computer. Most widely studied field in HCI is gesture recognition. Gesture recognition recognizes the current gesture of the user by detecting the user’s location and body parts using range sensors. Another popular form of HCI is the remote eye gaze tracking system. Remote eye gaze tracking is a field that makes use of the user’s eye to control computers. The current trend of research in the remote eye gaze estimation field is targeting on interaction with large TV.

Intelligent vehicleI is a field that develops safety devices using cameras to ensure the safety of drivers. Examples of such technology are the lane detection and the lane-departure warning technology which helps warn the driver of unintentional lane changes. Forward collision warning technology uses the camera to recognize the vehicle ahead and its location. It also helps to alarm the driver by calculating the time it takes to crash if there is danger of collision. Pedestrian detection and collision warning technology is a technology that recognizes pedestrian, its location and warns the driver when there is danger of pedestrian collisions by calculating the estimated time of collision. Autonomous vehicle technology is a field that integrates all of intelligent vehicle technology. It is a field that researches unmanned vehicle maneuvers by following road lanes while avoiding collision with obstacles, vehicles and pedestrian.

Research FieldsResearch Contents
Object Pattern Recognition
  • Face/Eye detection
  • Face recognition
  • Fast pupil detection for mobile devices
  • Measuring bone age
  • Various inspection devices
Intelligent Surveillance
  • People Counting
  • Tracking moving objects
  • Automated object classification
  • Intrusion detection
  • Loitering object detection
  • Unattended object detection 
Human Computer Interaction
  • Remote Eye gaze tracking
  • Gesture Recognition
Intelligent Vehicle
  • Forward collision warning system
  • Lane departure warning
  • Pedestrian detection and collision warning system
  • Autonomous vehicle