Tactical Perception Helmet
When I started this project, I envisioned a helmet that uses multiple cameras and microphones and as much sensors that I can fit on it with filtering and computer vision to allow the user to be more aware of their environment than if they weren't using the helmet. I was thinking of the helmet can be used in firefighting, underwater exploration, and for space exploration.
Physical Mask Construction
To start off, I decided to begin with the physical enclosure of the helmet, so there's something to attached the electronics to. I started printing sometime in 2018, I'm not completely sure since I started documenting everything I'm doing in Summer 2019. I started with the the part of the helmet that physically touches your head, the black part in the Figure 1, the main mask, the orange one, will be able to attach to. The first version of the mask is the white and black mask in Figure 1. The orange mask is the second version of the mask.
Figure 1: The head harness(black), Mask v1(black and white), Mask v2(orange)
I started to make the Mask v3 so that it can hold electronics, be easy to assemble, and be easier to 3D printer. The orange Mask v2 failed a couple times while printing, took a lot of support structures so it took a lot of filament overall. After I got the four parts to print, it was hard to align and JB weld together because each part was warped differently. Even after using clamps and giving the parts time to warp together, I wasn't able to align them properly during JB welding. To fix these issues as I broke up the mask into eighths and extra material with holes that align to the other parts to align them properly.
Figure 2.1: First iteration of Mask v3
Figure 3: Second iteration of Mask v3 with simulated colors
Figure 4.1: Third iteration of Mask v3, with alignment holes and extra material to add extra hardware
Figure 4.2: Third iteration of Mask v3, with broken up to make it easier to print
Figure 5.1: Close up of a part of third iteration of Mask v3
Figure 5.2: Close up of a part of third iteration of Mask v3
I am currently tuning my 3D printer to print these parts in ABS.
Audio and Visual on Raspberry Pi 4
During Summer 2019 I started the audio and visual part of of the helmet. I want to have electronic screens on the inside of the masks that have a HUD and feeds from one or multiple cameras on the mask, and have audio in and audio out. I'm looking for applications of computer vision and machine learning for the visuals, them I'm looking at filtering and voice command for the audio.
I currently am capable of having audio from one device, able to filter it, have it be able to send the microphone readings filtered to a speaker. I also have a stable camera stream, but capped at 720p 30fps with the Raspberry Pi camera.