Using a photo copy of R. E. Kalman's 1960 paper from an original of the ASME "Journal of Basic Engineering," March 1960 issue, I did my best to make an accurate version of this rather significant piece, in an up-to-date computer file format. For this I was able to choose page formatting and type font spacings that resulted in a document that is a close match to the original. (All pages start and stop at about the same point, for example; even most individual lines of text do.) I used a recent version of Word for Windows and a recent Hewlett Packard scanner with OCR (optical character recognition) software. The OCR software is very good on plain text, even distinguishing between italic versus regular characters quite reliably, but it does not do well with subscripts, superscripts, and special fonts, which were quite prevalent in the original paper. And I found there was no point in trying to work from the OCR results for equations. A lot of manual labor was involved.
Since I wanted to make a faithful reproduction of the original, I did not make any changes to correct (what I believed were) mistakes in it. For example, equation (32) has a P*(t)M(t) product that should be reversed, I think. I left this, and some other things that I thought were mistakes in the original, as is. (I didn't find very many other problems with the original.) There may, of course, be problems with my transcription. The plain text OCR results, which didn't require much editing, are pretty accurate I think. But the subscripts etc and the equations which I copied essentially manually, are suspect. I've reviewed the resulting document quite carefully, several times finding mistakes in what I did each time. The last time there were five, four cosmetic and one fairly inconsequential. There are probably more. I would be very pleased to know about it if any reader of this finds some of them. (John Lukesh's contact information.)
Please also copy Greg Welch (email@example.com) regarding any transcription errors.
Return to Welch and Bishop's Kalman filter page.
Return to Greg Welch's home page.