One-pixel Camera for Teaching and Research on Comparametric Equations and HDR

(You can download an SVG version (link) of the above drawing)

Let's build a camera with a sensor array that has just one pixel in it. This will force us to think about and truly understand what quantity it is that a camera can sense and measure.

Back in the 1980s and early 1990s I invented something called HDR (High Dynamic Range) imaging that is now widely used in smartphones and cameras and augmented reality vision systems, etc.. My general philosophy was to regard the camera as an array of light meters. The idea is to turn a camera into a scientific measuring instrument.

In this Instructable, you will build a one-pixel camera, and in so-doing, develop a deep understanding of what a camera is fundamentally, and what it can measure, once you see it quantimetrically (i.e. as a scientific measurement apparatus).

This is useful for teaching and research purposes, and makes it easy to do a variety of experiments with image sensing, lock-in amplifiers, scientific instrumentation, VR (Virtual Reality), AR (Augmented Reality), and RR (Real Reality, i.e. metavision).

The word "camera" is a Latin word that means "room".

In this context, "camera" is an abbreviation of "camera obscura" which means "dark room".

The original cameras were room-sized = big enough for one or more people to go inside the dark room and see pictures projected on a wall or surface, typically recorded by painting or drawing on the surface.

Let's begin with the optional step of making a zero pixel camera that simply displays the image it "sees".

You can start by creating a pinhole camera, and then adding a lens to make the image brighter (and sharper within a certain depth-of-field over which the lens is in focus).

In the example shown in the pictures above, we have a hole in the door to the lab (its on the 3rd floor gantry area so there's no security risk having the hole in the door) with black duct tape covering it, and a small hole to admit light into the lab.
The image can be viewed on the opposite wall, or it can be viewed on a piece of ground glass or plastic sheet stretched over a picture frame (if you have more time, take a large framed artwork, throw away the art, and buff up the glass with 400 grit powder, as indicated for example here (link).

Make a box and paint it black inside-and out. Or use an existing camera.

In one example, I used an existing camera and put a solar cell (monocrystalline) on a film holder and held it in place with tape, running the 2 wires out past the end of the film holder.

Choose an image sensor that is sized appropriately for the camera or box that you are using.

I also made one from a medium-format camera that normally takes 120 roll film. In this case I used a 2-inch square solar cell to nicely fill the 2+1/4 by 2+1/4 inch film image area.

In another example I made a small box camera in a small black cardboard box with a photoresistor in the box.

You can use just about any light sensor, such as a photocell, photodiode, solar cell, solar panel, or the like.

Try a variety of different kinds of photosensors, and make a few different kinds of 1 pixel cameras, and you'll begin to really understand what it means to sense light.

For the 1 pixel sensor you've likely chosen a photo resistor or a solar cell.

Connect the 1 pixel sensor to an indicator.

If you choose to use a photo resistor, connect it to a resistance or conductance meter (e.g. an "ohm meter").

If you choose a solar cell, connect it to a volt meter or ammeter.

Now you have a device that indicates a photoquantity of light.

The photoquantity, q, is neither radiometric (flat spectral response) nor photometric (response of the eye).

But you can still quantify it, as a quantimetric unit of measure.

To do this, consider ordered pairs of measurements at full strength and half strength, e.g. take one reading with the cell or light source half-covered, and one without, as illustrated in the drawing above, e.g. at four light strengths, (a), (b), (c), and (d), forming two examples of paired measurement, i.e. (a) and (b), followed by (c) and (d). Comparison of the measurement of a quantity with the measurement of half the quantity is called comparametric analysis.

See http://wearcam.org/photocell_experiment.pdf

In a previous Instructable, I taught how to display the sightfield or metavision of a camera (link).

Metavision is the vision of vision, in which you can see the extent to which a camera can see.

With the simple 1 pixel camera, you can explore the concept of metavision.

More simply, you can also just draw the camera's sightfield (link) with a pencil and paper, using a light bulb and a volt meter or ohm meter.

Try to develop a deep understanding of fundamentally what a camera measures.

Consider the light sensor outside the camera, and how it responds to light.

Begin in a dark room, with a single light source in which you can cover half the light.

Take a series of measurements of the light source in a particular position at both full and half strength, using black cardboard to cover half the light.

With a collection of ordered pairs, f(q) and f(2q), plot these as a scatterplot, and see if you can begin to understand the comparametric relationship between these ordered pairs.

Use the 1-pixel camera to try and learn as much as you can about comparametric equations.

Now you're ready to understand the fundamentals of HDR (High Dynamic Range) imaging.

If you build a 1 pixel camera and do anything interesting with it, or if it helps you understand the world of photography and imaging, please click "I made it" and upload some pictures or anything you learn from it.