This is a review of the ultrasonic camera, and my thoughts.
Problems and Solutions
1. It’s too slow.
On the current resolution, one picture takes five minutes. Increasing the resolutions also increases the pixels required to scan too. Also, the general distance to the target has an effect as well.
What can be done about this?
- Use two or more sensors
- Split the screen into sections, and let each sensor fill in its section. Stitch it all together later.
- Interference between sensors should be considered.
- Reduce time spent on distance measurement
- An ultrasonic wave is tied to a constant speed. An electromagnetic wave (like a real radar) should be considered.
2. The picture is warped.
The pixels are simply painted based on servo motor angles. This will squish down on the sides of the picture, creating a reverse fisheye lens effect. Currently this is a non-issue since there aren’t enough pixels to see that anyway. This will have to be considered if a improved high resolution & long range model is ever developed.
3. Irregularities
This seems to be a problem of the sensor itself. Some pictures show a dark pixel in the middle of bright ones. A better sensor will resolve this issue. Or by averaging values with surrounding pixels in the software side.
4. Severely limiting range
Another hard limit due to the speed of sound. This camera has a theoretical max range of 255cm, but the camera works best at 50cm distances. Further objects will be indistinguishable with the background white.
If this concept were to be taken any further, using electromagnetic waves will be necessary. Using multiple sensors and frequencies as the solution to problem 1 is already a thing in phase-array radar. I do want to try building this but I’m afraid the government or military will want to have a word with me. So it seems this is the end.