Locations

Artificial Eyes

 
LMDQ
Vietnamese-German University
Ho Chi Minh, Viet Nam
See full Bio Contact
 
Entry date: 20-May-2013
Final Submission: 01-Jul-2014
Artificial Eye is a sensor system whose purposes are to detect the shape of objects, estimate the distance and represent them graphically on a LCD…It will provide an advance data acquisition for a robot to interacting with the environment. Once a robot has a good understanding of a specific object, it would have a better accuracy when doing complicated actions such as handling objects, or moving around.
 

Project details

Motivation:

Many robots use infrared sensor or sonar to detect the obstacle or distance between the sensor and an objects. This is a primitive way for the robots to prevent collision or being stuck. By using sonar or infrared sensor, the feedback data is very limited and highly affected by the surface and the shape of the object. Therefore, the interactions between robots and objects are confined only to avoiding the object or finding path. For higher level of interaction, the robot must actually "see" the object, and the data gathered in this process are much more.

With that in mind, we want to build a system using 2 cameras and a processing unit based on Toradex Colibri T30 to have a better data acquisition of the environment around a robots. Our goal is that, by using data (images) collected from the 2 cameras, we can render a 3D object onto a screen with an acceptable approximation of the object's shape and dimensions.

Approach:

Hardware for the Artificial Eye are a mechanism structure on which 2 cameras are mounted and rotating for a small angle as analogous to human eyes, and the Toradex T30 as the data processing unit.

On Toradex T30, images from the 2 cameras are taken in real-time at a specific rate. Using digital signal processing, those images are transferred into a 3D representation of the object. Then the Artificial Eye will displays output on the LCD which are connected to Toradex T30.

Our expected result is to distinguish simple shape objects such as a sphere, a box, a pyramid or a cylinder... on a simple background. Then the robot can compare the distance and decide what is nearer. All that process will be display on a LCD.

Expected plan:

Step 1: Hardware built.

Step 2: Cameras connected and represent input image on LCD

Step 3: Complete a program to render a 3D object on LCD

Step 4: Finish DSP.

Step 5: Report

Comments Hide

Vaila Ruthvik
 

hi whatsup with your project???? did u get the kit ?did u start the thing ??? we are excited to know about your project?

krishna kumar
 

its cool......can u send me a abstract of ur project so tat t will be usefull for me to learn pls send as soon as possible....