Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

CMOS Sensor to check sharpness of focused image

Status
Not open for further replies.

Urahara

Member
Most of you should be familiar with a slide projector whereby a slide is projected onto the wall. If it is not focused, we move the lens up or down until we "see" a sharp image being projected onto the wall.

Instead of relying on visual inspection, I am thinking of placing a sensor on the wall which will check if the projected image is focused. Is there such a CMOS sensor that measures the light and then calculates whether the image is sharp enough? Sensor need not be colour sensitive (monochrome is fine) and needs to be interfaced to a microcontroller if possible.

Thks!
 
I do not know of a sensor like that.
If I had a digital camera looking at the pictire on the wall, or some part of the picture, then by looking at the data, some idea of focuse could be found. If the projector focus control was motor driven and could be driven through focus, back and forth, the best focus spot could be found.
Focus is when the blacks are most black, whites are brightest, and the edges are steepest. To say that another way, move away from gray.
 
Hello,

When focus is near you'll see the highest amplitude in the higher harmonics doing a two dimensional Fourier transform. That is, the higher harmonics should max out and most likely the lower harmonics will min out.
This of course assumes nothing is moving in the frame. Movement can be used to trigger a refocus.
 
Last edited:
Seems like such a sensor does not exist.

Think the only way to determine "focus-ness" is by doing an analysis (via software in the micro-controller) of the image data picked up by the sensor. If that's the case, anyone knows of commonly used algorithms/methods used for my purpose. Some code samples, etc will be useful.

Btw, have done some searching and it seems like the TSL1401 chip is a common one used by enthusiasts for their robotics vision projects.
 
tp://www.parallax.com/Store/Sensors/ColorLight/tabid/175/ProductID/566/List/0/Default.aspx?SortField=ProductName,ProductName
Should work.
 
Think like a manufacturer. Use a Sharp IR distance sensor to give actual distance of projector to surface, then auto-focus based on a lookup table.

Cameras have been using a similar process for decades.
 
Hi MrRB,

My web cam cant do it that way as it doesnt use enough power to shoot a beam 100 ft or so, and yet it does a pretty good job without shooting out any IR light.
My small hand held camera shoots out a visible light which can be seen with the naked eye and it illuminates a large center area of the pic. I assumed they were using harmonics to figure out the focus.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top