How do you turn a camera into a 'mouse'?

Jul 13, 2014 65 Replies

I didn't expect you to know what it means. I suggested that you find out what it means. Terms get bandied about so that you can go and find out what they mean.

Of course... but not as productive as when you follow up with searching, finding and reading some existing literature.

I'm glad you discovered that auto-correlation using X-OR doesn't work very well, but any intuition would have told you the same thing.

To handle image distortion, camera motion, seeing the same thing from slightly different angles, you need to do feature extraction and object recognition. A typical approach is to differentiate the image in two or more dimensions, pick the N highest peaks (or even better pick where a peak in one direction is a trough in the orthogonal direction, aka a corner) and then for each such point, make it into a triplet with its closest two neighbours. Figure out the angle and two side lengths of the triangle that forms, and index each triplet using those as lookup keys in an approximate lookup table. Now use the most prominent features to search in the same table made from the previous image, and you have a collection of probable match triplets. For each probably match, compute the transformation that moves the old triplet to overlay the new triplet. Out of all that, you'll get a massive cluster of similar transformations, and a lot of outliers. Discard all the outliers, and use a least-squares fit over the transformations in the cluster. That gives you a transformation from the perspective of the previous image to the current one with very high confidence. You can even use the systematic divergence in the errors to calculate what type of lens distortion is present, as Autostitch does.

But you could have learned all that for yourself the same way I did, by starting with Google.

Clifford Heath.

A UI-free image manipulation package implemented as several filters,

ms-dos or better

with a capital F

umop apisdn --- news://freenews.netfront.net/ - complaints: news@netfront.net ---

must be getting close I found this: NetPBM is a package with a myriad of programs that convert graphics from one format to another, and edit and analyze images. but no exact URL to download. did see mention of software on ubuntu! great I think that's the linus version next to replace all this !#$#!@$ MS stuff.

On a sunny day (Fri, 18 Jul 2014 06:41:46 -0700) it happened RobertMacy wrote in :

Use brain PLZ

formatting link

IF you have Linux (Linus is somebody), type apropos netpbm or apropos whatever_it_is_you_were_dreaming_of

After that exercise you could type: man netpbm

Strange I seem to have it installed...

Anyways it will not solve your problems I think. LOL But I admit I do not understand those, and suspect neither do you. LOL

Anyways, format conversion is normally done with man ImageMagick

type; apropos ImageMagick then man ImageMagick

it is a whole suite of programs, convert mogrify ....

well you can read.

But nothing to do with finding the position of something to 2.54 mm in a 200 x 200 feet area. So why bother.

I also know of a nice web browser, video player, audio player, video format converter, weather fax program, well a lot of programs, but that does not help you either. Linus has usually, by default if you install from some distro, some games too.

ah, text editor you need, news reader, web server, no does not help you either, How about a eeeh, calculator, octave, does not well gotta go, studying surface to air missiles.

again, google is NOT my friend. It takes me off to endless websites that have evidently paid for the right to get to me first, and a myriad of 'what sounds EXACTLY like what I want' expired websites, sigh. A simple search often takes over two hours and requires me to completely clean everything out of my system when finished. That, even with only judiciously going to a listed website, and even then often the link doesn't work, just hangs my system(s).

Thanks for the detailed description of 'feature extraction' Compression Labs, when they first started, was one of my clients. They taught me a great deal about image compression. I have yet to try some of those principles, like FFT to do some of the processing - that's fast. position shift is phase shift and distance is merely frequency ordinate compression, but with so few pixels don't feel real comfortable doing that type of effort, especially when the simple minded shift/comparison works so well.

I did find out that the XOR works pretty well, but eats a huge amount of calculation since the matrices are so large. I found that doing linear matrix reduction, which essentially creates a gray scale from the B/W image, does make for small matrices, but wasn't sure how to handle the image comparison. Turned out that using the 'hanning shaped' window, inverse cosine shifted to be 1 at max and zero at min; actually enhanced the 'contrast' for the comparisons, compared to my simpleminded 'window approach', or more like bin approach.

Now if I find some software doing what I'm doing I'll be able to trust it and understand its quirks.

thanks for the URL to the manual! from there may be able to get the other stuff without getting overloaded with tracking cookies.

Actually the scaled compariosn to the 'ball mouse' performance is around

10 feet by 10 feet with resolution of 100 mils. From the experiments done here seems doable to not have excessive 'slippage' over that range, but will have a lot at 200 feet but don't care, since distortion is acceptable.

yes meant linux.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required