The benefits of scanning in Greyscale
Inherently the quality of microfilm images varies considerably. It is not un-common to find images of differing density within a microfiche, particularly where a fiche contains 16mm and 35mm images, have been filmed on different cameras or updated, especially when these have been produced over a number of years with text of varying colours and quality.
Whenever a document or microfilm scanner digitises a document to a black and white format it is in fact scanning the document as a greyscale image. Software then performs a process called thresholding to convert the image to black and white (mono) before saving it to disk.
During the thresholding process the software evaluates individual pixels (picture elements) and determines from the greyscale value whether the pixel should be switched on (converted to a black pixel) or switched off (converted to a white pixel). The quality of the software algorithm to convert the greyscale image to black and white is very important as it will throw away information which cannot be retrieved!
The image below illustrates the resultant image of the thresholding process.
The darker areas which appear in the greyscale image have been converted to black whilst the lighter areas are converted to white during thresholding. The technical term for the number of colours in a digital image is known as colour depth, the colour depth of an image is denoted by the number of bits (binary digits) that specify the colour of the pixel i.e. 8 bit or 16 bit.
Alternatively you may see greyscale referred to by the number of greyscale levels in the image i.e. 16 or 256 levels, the greater the number of levels or bits corresponds to a greater amount of information
For a black and white image we require 1 bit of information to describe the pixel value either 0 (black) or 1(white) whereas in a 16 bit greyscale image we can describe a pixel with 256 levels of grey from white to black.
The benefit of scanning photographs in greyscale is obvious, the value of scanning textual documents is often not so obvious. We tend to assume all textual documents are generally black (or dark) ink on a white background however, this is not always the case as with documents of poor quality, poor lighting (i.e. when microfilmed) or where dark text has been placed on a shaded background.
Image “A” below demonstrates a greyscale document which has been converted to black and white resulting in the loss of information and consequently illegible text in the bottom portions of the image.
When compared to image “B”, the benefits for scanning and saving images in greyscale are obvious. It can be seen we have retained as much information as we can from the original document. By utilising our post scanning thresholding software we can control the quality of the resultant black and white image. If we are using optical character recognition to create a searchable document we can carefully control the resulting image presented to the OCR engine.
Even when converting images to black and white to use as working documents, Datastor provide clients with a fully indexed copy of scanned data in greyscale at no extra cost, giving them the opportunity to refer back to the original greyscale image (and enhance if necessary) to produce a more legible image or print for reference as required.