Key Benefits of ViewONE - Anti-Aliasing

Anti-aliasing for viewing clarity – a crucial element of effective web-based document sharing

Web-based image viewers that allow the secure sharing and annotation of documents are a hugely useful and important tool for a wide range of businesses. Even when users are spread across the remotest of locations web-based viewers can help to improve productivity and provide a framework for effective document control. However, all the clever functionality and sophisticated tools are of little use if the documents to be worked on are of a poor, difficult to view quality.

Anti-aliasing is a system that can significantly improve the viewing clarity of even the poorest quality viewed images and document scans.

Looking behind the growth of web-based image viewing

The Internet has changed the way we live our private and work lives beyond recognition over the last 10 or 15 years. Tim Berners-Lee’s great little idea that was hatched on Christmas Day in 1990, with the first communication between an HTTP client and server via the Internet, has also been the catalyst for a huge number of other innovations. One of these is web based image viewers that help organisations of all sizes work on scanned documents and other image files, even when users are spread across the globe.

The speed, ease and control that these viewers allow in delivering viewing and annotation capabilities for scanned documentation to an entire user base are impressive. This is especially true when they are compared with outmoded approaches such as helper applications (i.e. external viewers to the browser) and those requiring pre-installation or specific IT skills and permissions. These typically old applications have attempted to adapt rather than redesign for the new world. Such applications also often require users to work outside of the existing browser framework and so result in delays and additional costs. Furthermore, they usually do not integrate effectively enough into the current application to add value and at times actually remove value from the existing work practice.

The benefits of web based image viewers that have been designed from scratch specifically for the web can translate to an almost unlimited range of industry sectors. From accountancy to fashion, education to automotive, financial institutions to engineering establishments and health service companies, in-fact any organisation needing to share and annotate documents, diagrams and other visual information can reap benefits from the technology.

The Internet has created a demand for immediacy of communication and information sharing in the workplace. Text, instant messaging and email have fuelled and met this demand for the written word. Meanwhile web based image viewers such as Daeja's server licensed ViewONE Pro & ViewONE products help ensure that image sharing and collaborative document manipulation keep up with the blistering pace and expectations of the online revolution.

Why do we need anti-aliasing?

While it is possible to create images of a resolution that should satisfy the viewing needs of even the most demanding applications, the availability of a viewing medium – i.e. computer monitors – with the ability to display such detailed images at their full resolution is both a technical and economic challenge. This is largely due to the difficulties in obtaining high density and accurate colour pixels on phosphorous monitor tubes. More recently this problem has been seen with the high cost of producing similar high density LCD panel monitors.

In order to display a high resolution image on a lower resolution screen, pixels need to be ‘thrown away’; this naturally leads to a degraded image being presented to the user. Further problems occur when, for example, reducing a 200dpi image file to 72dpi for display purposes results in having to throw away one in every 2.78 pixels. This is clearly not possible so either one in two or one in three must be thrown away.

Where fewer than the desired number of pixels are thrown away the result is more pixels than are needed and the image is too large, and where more than the desired number are thrown away we end up with an image that is too small.

The problem of erroneous pixels known as ‘aliasing’ occurs when, in order to maintain the correct length and width of the original image, the number of pixels thrown away across and down the image is varied.

Also the removal of any pixels can only result in loss of some, and in some cases, a lot of clarity. Without employing the technique of 'anti-aliasing' this results in users having to repeatedly 'zoom-in' to be able to 'read' the image properly. The process of 'zooming-in' allows those missing pixels to be re-introduced, however, it also can add inconvenience at best and huge delays at worst to what should be a relatively simple task of just viewing an image.

What is anti-aliasing?

The solution to the problem of aliasing and missing pixels is named ‘anti-aliasing’. This is a term which can apply to both audio and visual signals to describe the representation of a high-resolution signal at a lower resolution. In the case of displayed images, this technique involves applying a complex formula to ascertain and execute the in-filling between the lighter and darker pixels of an image with grey or mixed colours and converting alias pixels to grey. This effectively blurs or smoothes the edge between light and dark areas of an image. When viewing this type of conversion from a distance, the eye sees a more consistent and clearer joined-up image as opposed to the previously pixillated or broken up version. In effect, it's fooling the eye, however it is an extremely productive way to help improve the readability of an image as the brain essentially makes up for the missing pixel data by subconsciously 'filling in the blanks' as a user reads images presented in this way.

The fundamental problem to be tackled when developing anti-aliasing techniques is to optimise color changes whilst keeping the time to do this process to a minimum. Also, and not to do it so much as to add too-much grey so that the images appear blurred which can result in worse effects, perhaps even eye strain. The time issue is well illustrated by considering that a typical image may have in excess of three million pixels and to perform even simple anti-aliasing may take several times this number of calculations. The duration of an anti-aliasing process may therefore take several seconds, or perhaps even minutes – difficult to accept in our speed driven world!

To address this issue, companies such as Daeja Image Systems has developed various anti-aliasing techniques to maximise speed whilst still outputting an image of good quality. The type, size, and colour of a given image will lead to one approach emerging as the best route to achieving viewing clarity for that file.

Horses for courses – different anti –aliasing techniques

For monochrome, black & white or colour, text, photography or diagrams may all have greater clarity depending on which anti-aliasing approach is used. From the users point-of-view the most important aspect of any anti-aliasing technique is the trade-off between processing speed and viewed image clarity.

The first technique to be introduced by Daeja Image Systems, and one that is still relevant for many monochrome image types, is ‘simple-averaging’. However, as the diversity of image types grew over time, and in response to an issue whereby user’s eyes could become tired or strained when looking at simple-averaged images on very low quality monitors, a more complex technique known as weighted-averaging was introduced. The algorithm used in this type of technique performs a biased averaging calculation based on the fairly sound assumption that text, or the lines of a diagram, are usually darker than their background.

Approaches such as those described above work well for most monochrome text based images, however when viewing very large line-based drawings such as machinery plans or perhaps huge automotive plotter plans that can measure in excess of 10 metres long when printed, an apparent fading of the lines can occur. In these applications it is useful to employ a different approach, that rather than averaging out segments of an image and using grey pixels, ensures that if any pixel in that segment is found to be black then that segment is always displayed as black. This is a very effective way of showing fine detail in localised areas of a large image. Referred to as ‘scale-to-black’, this method can be offered along with the other monochrome image techniques described in a single software package thus allowing users to ‘toggle’ between techniques to see which gives the best combination of clarity and speed for a given image.

Letting the user choose - anti-aliasing for colour images

Anti-aliasing of colour images, as you might, expect is a more complex challenge than is the case with black and white. On occasions it may improve the display quality of the image and on others it may have a negative effect causing blurring. Factors that can influence whether this is the case include the original image quality, preferences, quality of the computer monitor and even the focus and colour ability of the users own eyes. For this reason it is best to provide the user with the option to ‘turn on’ or turn off’ anti-aliasing for colour images.

The benefits of anti-aliasing techniques used on black and white images do not read across to colour applications. For example, if weighted-averaging is applied to a colour image, then the process of performing the calculations on multiple colours can have such a big impact on speed and performance that any potential benefits on resultant image quality are outweighed. Scale-to-black is even more clearly defined as a process purely for black and white images.

Simple-averaging emerges as the best approach and compromise for anti-aliasing of colour images. Averaging calculations must be performed separately on each colour pane and this amount of processing can take time, so once again a careful balance between quality and speed needs to be achieved. Because of this, and also because of the uniqueness of each and every image, providing users with a fast easy way to see and choose between an unprocessed and a processed image is key to surviving in a fast faced, dynamic workplace.