Colour to grey scale and related transforms

每年都要制作数百万张灰度差的彩色图像复制品。通过我们的研究，我们的目标是制作出数百万张好的灰阶照片。使用彩色数码相机捕捉图像并将其显示在彩色显示器或使用彩色喷墨打印机上，现在越来越普遍。然而，仍有许多场合彩色图像是用灰度再现的。使用黑白打印机、复印机和传真机仍然是每天都在发生的事情。不幸的是，将彩色图像转换为灰度图像，无论是通过这些设备，还是通过其他方式，往往导致图像的再现效果很差。我们研究的目的是通过仔细考虑我们自己的视觉系统的局限性，并利用彩色图像形成的潜在物理原理，开发一种方法来获得彩色图像的最佳灰度再现。通常，给定图像点的颜色由三个数字编码，在灰度变换中，这三个数字被简化为一个灰度值。更一般地说，颜色信息可能使用N（其中N>3）个数字进行编码，在这种情况下，能够推导出将这些N个数字减少到3或更少的方法是有用的，这样记录的N维信息就可以在传统成像技术上显示。我们将扩展我们的研究，也考虑这些情况。我们的研究将改善灰度设备上的图像再现（例如，改进复印和传真），并为我们提供更好地可视化所包含的信息的能力，例如卫星图像。我们还期望有可能利用我们开发的技术来制作彩色图像，使色盲观察者能够准确无误地观看彩色图像。此外，这项工作将改进现有的图像处理算法，并更好地理解我们自己的视觉系统是如何感知颜色和亮度信息的。

Every year, millions of poor grey scale reproductions of colour images are made. Through our research we aim to make millions more good grey scale prints.It is now increasingly common to capture images using colour digital cameras and display them on colour monitors or using colour inkjet printers. However, there are still many occasions when colour images are reproduced in greyscale. The use of black and white printers, photocopiers and fax machines is still an every day occurrence. Unfortunately, the conversion of colour images to their greyscale equivalent, either by these devices, or through other means, often results in a poor reproduction of the images. The aim of our research is to develop a method to derive the best possible greyscale reproduction of a colour image, by a careful consideration of the limitations of our own visual system, and by exploiting the underlying physics of colour image formation. Usually, the colour at a given image point is coded by three numbers, and in the greyscale transformation these three numbers are reduced to just one grey value. More generally, colour information might be coded using N (where N>3) numbers, and in this case it is useful to be able to derive methods to reduce these N numbers to 3, or fewer, so that the N-dimensional information recorded can be displayed on conventional imaging technology. We will extend our research to also consider such cases. Our research will lead to the improved reproduction of images on greyscale devices (e.g. improved photocopying and faxing) as well as providing us with the ability to better visualise the information contained in, for example, satellite images. We also expect that it will be possible to exploit the technology we develop to make colour images that can be viewed without error or confusion, by colour-blind observers. In addition, the work will lead to improvements in existing image processing algorithms, and to a better understanding of how our own visual system perceives colour and brightness information.