MULTIPIXEL SCANNING SYTEM
The goal of this project was to build an imaging system. The physical components of this system included a pico projector, ambient light sensor, some passive electronics, and a TI MSP430 Launchpad microcontroller. The completed setup of the imaging system looked like this:
Here is the system flow:
-
The ambient light sensor converts light energy into electric current; more ambient light leads to more electric current.
-
The current produced by the ambient light sensor then travels through passive electrical components, which translates the current into a detectable signal (a voltage signal).
-
The TI Launchpad reads this signal, and after some linear algebra the voltage signal gets converted into an image.
-
Further improvements (like modifying the background light) are made via software to increase the quality of the image.
​
Background Knowledge & Process:
Digital images are composed of a grid of pixels, each of which represents a single color. To create a digital representation, optical imaging systems measure the energy, equivalent to light in this case, reflected off of objects to determine the values of these pixels. Because different colors reflect different amounts of light, measurements of reflected light intensity can be used to differentiate between colors.
Commercial cameras, such as those found in smartphones, use chips called CMOS image sensors to measure millions of pixel values at a time (think of this as millions of devices akin to the solar cell that all fit on an approximately 30mm chip). In contrast, our system could only measure one light intensity value at a time, so we instead used a controlled energy source, the projector, to illuminate specific regions of the image one region at a time. We then used the solar cell to record the energy that reflects off of each region. The solar cell converted light to induce a flow of electrons, and the circuit then output a voltage corresponding to light intensity. The specific configuration of the circuit set the solar cell up so it was most sensitive to changes in light.
This changing voltage was a set of real numbers and much closer to something we could use to represent our image on a computer. We used a microcontroller (Launchpad) to take samples of the signal and converted it to a set of numbers we could then use in IPython.
