Embedded Files
Scope of our research interest and project is wide from Robotics application to Computer vision. You can see our past and current project in table below. Clicking project name will direct you to each own page and you might see our interesting demo video.
Ongoing Projects
Humanoids - "HUBO"
Goal: Develop Bipedal walking Humanoid robot
Features: 1. Cycloid Algorithm. 2. Use Hubo of Kaist. 3. Controlled by ACES conductor. 4. Simulated by Webots.
Extract 3D information of Object
Goal: Getting 3D information of Objects ( 6DOF)
Features: 1. SIFT, BBF and RANSAC baseD. 2. Getting 3D information (Location(Including Depth) and Rotation) of Target. 3. Epipolar Gemetry using Stereo Webcam
Motion Capture System - "MOCAP"
Goal: Develop Ultrasonic Motion capture system and augment 3D augmented Virtual Reality
Features: 1. Motion Capture system using Ultrasound Receiver and Transmitters. 2. SURF based target detection. 3. 3d augmented Virtual Reality using OpenGl
Goal: Develop a Modular Robot which has Cognitive features
Features: 1. Modular Design: every parts of robot is modular so can be changed easily. 2. This robot can express its emotion and is always motivated to do something which is relevant to stimuli from environment. 3. This robot can remember past event with people and use the memory at next meeting 4. Main controller is Fitpc2 and parts are manufactured by using 3D- printer
Floating Autonomous Robot - "UP"
Goal: Small sized Region Exploring Robot
Features: 1. Autonomous robot which is controlled by Microcontroller. And IR sensor provides obstacle detection during flight. 2. Inspired from Pixar movie "UP". 3. Sparkfun " PIC24HJ64 Board" is used
Goal: Given an agent with many alternative for action, we can give robot the bias toward certain classes of action under varying conditions reveals the personality of the agent
Features: 1. Probabilistic in nature. 2. Incorporates Personality Generator. 3. Behavior Selection based on Personality
Past Projects
Sprawled Edubot ( Rhex class ) : This project was conducted at Penn under the FIBR (Frontiers in Biological Research) project framework.
For More Details : Kod*Lab (http://kodlab.seas.upenn.edu/)
Goal: Building SprawlHex and Anlysis locomotion charasteristics.
Features: We present experimental data on the dynamics of Sprawl-
Hex — a hexapod robot with adjustable body sprawl consisting of time trajectory of full body configuration and single leg ground reaction forces. The dynamics of SprawlHex is compared and contrasted to that of insects.
SprawlHex dynamics has qualitative similarities to that of insects in both sagittal and horizontal plane. SprawlHex presents a step towards construction of an effective physical model to study arthropod locomotion.
CogbotGoal: Build a Autonomous Cognitive Robot.
Features: 1. Autonomous robot which is controlled by Fitpc2.
2. 2 Logitech Webcams provide face detection and obstacle detection.
3. Cogbot can recognize and generate speech.
4. Cogbot has 4 Emotions like Happy, Sad, Angry and Scared
5. Also has 5 Motivation like Navigate, Charge Battery, Conversate with people. Make a advice to people and Play with other Robot.
Directional Loudspeaker Goal: Build a Directional Loudspeaker.
Features: Audible sound from this speaker is highly directional and convergent, which means you have to be in the "right spot" to hear the sound!
As demoed in the our video, only when the speaker face right directly to you that you can hear the song, and if the speaker is rotated a slight angle, the audience can barely hear it.
Autonomous Robot in Biosphere2
Goal: Build Small Sized UGV(Unmanned Ground Vehicle) Robot.
Features: 1. Autonomous robot which is controlled by Q1.
2. Logitech Webcam provides obstacle Terrain Segmentation.
3. A Howard's terrain analysis method fr traversability is used.
4. Dynamixel Motor is used for Steering.
5. Also can controlled manually by Controller window in website.
Goal: Build Multiple Target tracking system.
Features: 1. Multiple Target tracking system using Computer Vision
2. Sponsored by ISCA technology
3. Developed for detecting flying bees for research purpose
4. Algorithm was Implemented by Fitpc2 (considering its good portability) since bee tracking should be done in outdoor environment
Object Recognition and Optical Flow
Goal: Recognize various kinds of Objects and detect motion between consecutive frames.
A. Object Recognition
Features: 1. David Lowe's SIFT - Scale Invariant Feature Transformation
2. Bartlett's SEEMORE - 102 Features Histogram
B. Optical Flow
Features: 1. SIFT based Optical Flow.
2. Lucas-Kennedy using Shi-Tomas feature points
Image Analysis for Unmanned Vehicle
Goal: Image Analysis for Unmanned Ground Vehicle
Features: 1. Using Be Franklin Teams' Gaussian Approach
2. Implemented by Matlab
3. This algorithm is adopted by our small sized UGV with A Howard's approach
Goal: Building Small sized Segway Robot
Features: 1. Self balancing using PID controller
2. Equipped with Gyro and 2 Dynamixels
Dynamical Analysis in Leg of HumanoidsGoal: Robotic Biarticulate Muscle Leg Model SimulationFeatures: 1. Analysis of Biarticulate Muscle leg in Humanoid robot. 2. Dynamics study using Dap3d 3. Run by Matlab
What Makes Us Different
We have many backgrounds and experiences in various area which cover Cognitive Robotics, Ultrasound system, Computer Vision, Unmanned Autonomous Robot and much more. Our past and current projects which are listed above prove the fact. And we are strongly motivated to work and to study in those area.
Page updated
Google Sites
Report abuse