Final Design

Summary of Final Design 

Hardware Setup: 

                                                                         Experiment Station Construction 

Figure 1: Overall Hardware Setup 

Figure 5: Hardware Setup for High Speed Camera and Galvo

Frames: 

Figure 6: Hardware Setup for Low Speed Cameras and Frame

    The framework would be able for us to install all necessary hardware components including cameras, galvos and power supply. Also, the framework should have some degree of flexibility. In the future conduction of experiment, there may be slight change of positions of cameras and galvo system for adjusting the accuracy. The final decision for constructing the framework is to build it with a combination of T-slots and breadboard. Because our arena is expected to be a 20*20*20 cm cubic box, then the final decision for the frame is a 1218*1967*800 mm rectangular box constructed using aluminum double T slots and is mounted on the fixing table. Then, one additional double T-slot would be fixed on the top face to mount the high-speed camera.

Installment of Low Speed Cameras: 

Figure 7: Low Speed Camera and Mini Ball head

    We have four low speed cameras and we decided to use two of them for triangulation. These two cameras should record the whole arena. We need to make sure that every spot in the arena should be covered by at least two low speed cameras, so we can do triangulation according to data recorded by at least two cameras then we could calculate the coordinate of our target. Then the final decision is that four low speed cameras have been mounted using a mini ball head on the top four corners.

Installment of High Speed Camera and Galvo: 

d = 780 mm    l = 200 mm    h = 130 mm  

Ø: scan angle= arctan(l /(2d)) = 7.31 degrees 

r : hole radius = h*tan(Ø) = 17 mm

Actual radius: 34 mm, FS = 2

Calculation Needed for Installing Gavlo

    The high speed camera is used for recording the video of our target and the galvo system is used for reflecting our target into the high speed camera. With an inappropriate galvo position, the high speed camera will either not be able to see the entire arena or see the edges of the breadboard because of the limitation of angles of mirrors. Our final decision is to drill a see-through hole with a radius of 34mm center on the breadboard based on the range of the galvo. The high speed camera could be mounted on a rail and two linear stages with micrometer. Therefore, the position of high speed camera could be adjusted really accurately on the x and y directions. Then the galvo system is mounted on a linear stage that could move on the z direction. Therefore, the center of the galvo could be adjusted accurately according to the height of high speed camera. Also, because the change of position of galvo on the z direction will not influence the calibration of other cameras, this design is more convenient for future experiments.

Figure 8: High Speed Camera CAD Assembly     Figure 8: Galvo CAD Assembly                                         Figure 9: Galvo Assembly

Software Coding:

For coding, we need to develop some C++ codes to ensure that all cameras and galvo are correctly calibrated and construct the data flow among the low-speed cameras, the computer, the high-speed cameras and the galvo, as shown in the flowchart below.

All codings are present in code files.

And the rotation of Galvo mirrors and the view of High-speed camera are shown in two GIFs below:

            Demo for Galvo Rotation                                                                                                 Demo for Galvo Rotation 

Performance

Two videos for tracking at z=0 and z not equal to zero are attached in the multimedia part. After doing the LED testing, we calculated the distance between the center of view and the LED light for representing the accuracy. And the summary of performance is listed below. 

• The galvo always follow the LED as long as the height of LED is within our constrained space.

• The deviation from center of high speed cameras is approximately 0.3 millimeter.

• The latency at 33-50 Hz is negligible. However, due to hardware limitation, frequency higher than this range will start to create significant latency. 

Figure 4: Time v.s Dis between Center and LED (pixels)

Table 1: Statistics of LED Testing

Future Improvement

Table 2: Potential Future Improvements