Browse Prior Art Database

Intelligent Scaling of Fonts in Annotations

IP.com Disclosure Number: IPCOM000028964D
Publication Date: 2004-Jun-09
Document File: 2 page(s) / 32K

Publishing Venue

The IP.com Prior Art Database

Abstract

Any image when displayed or printed needs to be annotated with useful information. For example, all clinical images need to be annotated with at the least the patient name, patient ID and other parameters useful for identification of the image and aid in diagnosis. One method of annotating these images is to burn in the annotation into the image itself before they are sent to the printer or to a display device. In order for the images to be legible, invariably the images are zoomed by the printer and printed (or by the display device). . If a fixed font is selected for rendering the annotations, then they get scaled differently as the images get scaled differently (since the annotations are a part of the image itself). The rendering engine on the host needs to be more intelligent to use different size annotations on different images so that the annotations are legible when printed.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 52% of the total text.

INVENTION TITLE: 

Intelligent Scaling of Fonts in Annotations

PROBLEM/BACKGROUND: 

Any image when displayed or printed needs to be annotated with useful information. For example, all clinical images need to be annotated with at the least the patient name, patient ID and other parameters useful for identification of the image and aid in diagnosis. One method of annotating these images is to burn in the annotation into the image itself before they are sent to the printer or to a display device. In order for the images to be legible, invariably the images are zoomed by the printer and printed (or by the display device). The zoom factor by which these images get zoomed depends on the size of the image and also the view port size. For example, a CR image (1760X2140) needs to be zoomed differently than an MR (256X256) or a CT (512X512) image. If a fixed font is selected for rendering the annotations, then they get scaled differently as the images get scaled differently (since the annotations are a part of the image itself). For example, if the selected font renders the annotations on a CT image to be clear and legible, the same font if used with a CR image will produce annotations that are very small and may not legible at all. This causes inconsistency and in some instances the annotations may not be legible on one image when it is legible on another. The rendering engine on the host needs to be more intelligent to use different size annotations on different images so that the annotations are legible when printed. 

INVENTION DESCRIPTION: 

This invention describes an algorithm, which enables the rendering module to select a suitable font from a pool of fixed size fonts depending on the image size and the view port size (where this image will be rendered). 

At a high level the algorithm selects a font size that depends on the image size and the view port. The algorithm avoids such situations where smaller images (such as MR 256x256) might end up having too big a font (and hence covering the interesting part of the image itself, like the anatomical details) or having a very small font for a big image like a CR (making the annotations illegible).

As an example, let us consider an MR image of say 256x256 dimension, being printed on to a film of size 8”X10” size. The film resolution would be (1000X1500). In order to fit the image into the view port the printer would zoom the image by a factor of » 3.9. Simultaneously annotations will also be zoomed by the same factor. Now, let’s consider a CR image (1760X2140) to be printed on the same film. This time the printer will scale the image by a factor of » 0.6.Clearly, if the same font size is used in both cases, the annotations will look big in the case of an MR image in comparison with that in the CR image case.

A solution to this problem is to sca...