Actually, extracting a 3d image is not required. The patterns that human eyes percieve as 3d images can easily be percieved directly by machines. Sources for making SIRDS images from text are easily adapted as sources for reading them as text. Bear On Mon, 4 Dec 2000 mmotyka@lsil.com wrote:

I've got sources kicking around somewhere that do this with noise patterns in monochrome bitmaps. I'm not sure if the effect using ascii is achieved in the exact same way - no doubt it is very closely related but not as dense as the bitmaps.

Mike

...

Well, not crypto but stego.

I was looking for ways to use existing infrastructure for transporting ascii messages to send info which is not harvestable by machines.

This came up:

It is intended for cross-eyed free viewing. Cross your eyes until the Vs and Xs at the top and bottom overlap with the adjacent ones.

V V V V V OIWEQPOISDFBKJFOIWEQPOISDFBKJFOIWEQPOISDFBKJFOIWEQPOISDFBKJF EDGHOUIEROUIYWEVDGHOXUIEROIYWEVDGHEOXUIEOIYWEVDGHEOXUIEOIYWE KJBSVDBOIWERTBAKJBSVEDBOIWRTBAKJBSOVEDBOWRTBAKJBSOVEDBOWRTBA SFDHNWECTBYUVRGSFDHNYWECTBUVRGSFDHCNYWECBUVRGSFDHCNYWECBUVRG HNOWFHLSFDGWVRGHNOWFGHLSFDWVRGHNOWSFGHLSDWVRGHNLOWSFGLSDWVRG YPOWVXTNWFECHRGYPOWVEXTNWFCHRGYPOWNVEXTNFCHRGYPWOWNVETNFCHRG SVYUWXRGTWVETUISVYUWVXRGTWVETUISVYUWVXRGWVETUISVYUWVXRGWVETU WVERBYOIAWEYUIVWVERBEYOIAWEYUIVWVERBEYOIWEYUIVWLVERBEOIWEYUI EUIOETOUINWEBYOEUIOEWTOUINWEBYOEUIOEWTOUNWEBYOETUIOEWOUNWEBY WFVEWVETN9PUW4TWFVEWPVETN9UW4TWFVETWPVET9UW4TWFBVETWPET9UW4T NOUWQERFECHIBYWNOUWQXERFECIBYWNOUWFQXERFCIBYWNOFUWFQXRFCIBYW VEHWETUQECRFVEVEHWERTUQECFVEVEHWQERTUQCFVEVEOHWQERUQCFVE UIWTUIRTWUYWQCRUIWTUYIRTWUWQCRUIWTXUYIRTUWQCRUIBWTXUYRTUWQCR IYPOWOXNPWTHIECIYPOWTOXNPWHIECIYPONWTOXNWHIECIYLPONWTXNWHIEC R9UHWVETPUNRQYBR9UHWVETPUNRQYBR9UHWVETPUNRQYBR9UHWVETPUNRQYB

Any sources for making these ?

I know that it can be machine-read, by extracting 3D figure and OCR-ing, but not today easily and in quantity.