Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!uunet!mcsun!ukc!stl!crosfield!rak From: rak@crosfield.co.uk (Richard Kirk) Newsgroups: comp.graphics Subject: Re: Can handheld scanners scan BACKLIT images? Message-ID: <10397@suns501.crosfield.co.uk> Date: 25 Jun 91 10:47:54 GMT References: <2835@umriscc.isc.umr.edu> Reply-To: rak@crosfield.co.uk (Richard Kirk) Followup-To: comp.graphics Organization: Crosfield Electronics, Hemel Hempstead, United Kingdom. Lines: 53 > The internal light may be sufficient, if you back the transparency > with a solid white ground -- eg, oversize white paper. It all depends on the contrast range on your X-ray. X-ray film can have useful information at densities beyond 6. An 8-bit CCD hand-held scanner with linear A/D will only go up to about 2. A sensible radiographer will adjust the doses and add anti-scatter shielding, but sometimes that cannot be helped when looking at the edge of a cylinder, etc. The usual problem is too little dose. Whether it will work often depends on... What is the biggest density in the region of interest? > Just because > *you* view Xrays on a light-box doesn't mean the scanner is restricted > to viewing the Xray negative in that position. The light goes twice through the film, so the effective density is roughly doubled ( the exposure on high sensitivity film is visibly mottled ). If the original density range was less than 1 then you might get away with it. You will also have a contribution from the backscattered light if you cannot block off the light source. This signal will vary with the smoothness of the film which in turn can depend on the condition of your fixer, and the agitation in your bath. I doubt if it is significant under normal conditions, but beware of reading too much into small structures in the shadows. >I don't know if Xrays have the image on the 'front' or 'back' of the carrier. On all X-ray films I have dealt with there is an emulsion on both sides. This doubles the sensitivity. Under normal viewing conditions the parallax error between front and back will be small. Is this still true with a hand-held scanner? If you have too much contrast on your film for the details to br readable then you can halve the density by bleaching one side of the film with cotton wool and ammonia. A less final method would be to print the image onto a soft paper. If you have the facilities I would recommend scanning a print of an X-ray, if only to establish that scanning the x-ray directly works. I would expect the print to have all the detail in the original with a better contrast range, perhaps a better magnification, and less parallax errors (if significant). If you are trying to get quantitative dose measurements of the film rather than just detect edges there would be an extra round of chemistry to keep constant, but that shouldn't be too hard. Please let us know if it works! -- Richard Kirk Image Processing Dept Crosfield Electronics Ltd. U.K. 0442-230000 x3361/3591 Hemel Hempstead, Herts, HP2 7RH