In order to distribute creative image content, authors go to great lengths to safeguard that it is used and reproduced in the way they want. In this paper we propose to use the framework of ICC color management in a novel way, in order to provide a means for determining two aspects of image reproduction: 1. on what devices and media it can be printed and 2. who authors want their content to print. We achieve this by creating a custom pair of ICC profiles, one part of which acts as a “public-key” and the other as a “private-key”, while the color management engine acts as the encoder/decoder. The key to this approach is to depart from the pre-requisite of a common Profile Connection Space and instead generate a multitude of encrypted spaces. If the profile used to encode an image uses the same encrypted space as the profile used to decode it, the image is reconstructed without error, if this is not the case the reconstructed image is unusable as the colors are scrambled. The main benefit of this approach is that it requires no changes to the typical workflow of converting between color spaces using existing software, while affording control over content in a safe and easy way. Introduction With more and more image content being distributed over the internet by means of cloud–based solutions (such as www.snapfish.com or www.shutterfly.com), the problem of creative content management is becoming increasingly important. Creative customers such as professional photographers or fine artist are known to be very keen on keeping control over the way their content is presented and reproduced. In the words of Magnum photographer Guergui Pinkhassov: “Sometimes I have not even recognized my own photographs. I have even hesitated to call them my own. [...] Whoever controls the editing of a photographer, controls his fate.” [1] (note that by ‘editing’ the process of developing and printing a photograph is referred to here). In this paper we focus specifically on the problem of how images are handled in the printing context and present a method that gives some control to the author over how their content is reproduced, specifically, what device/paper it is printed/viewed on and what rendering intent is employed. There are numerous approaches that address the broader topic of safeguarding creative content, for example Nikon’s Image Authentication Software [2] that enables detecting the alteration of images (for high-end Nikon DSLRs only), or digital watermarking solutions such as Digimarc [3], shipping with Adobe Photoshop as a plug-in, that embed information in the image for author identification or copyright purposes. Such solutions however, do not prevent uncontrolled reproduction and in the worst case can even be content destructive. Another alternative used is to add visual watermarking with a copyright message, but this is counterproductive in the case of distributing image content with the aim to enable a determined type of reproduction. Our approach instead is one that is non-destructive and focuses on key aspects that can be chosen at the point of printing an image. Furthermore, it is straightforward to use, as it doesn’t require custom tools to be employed by the end-user. We achieve this by departing from a key principle of ICC Color Management in a controlled way, by creating a custom pair of profiles that give desired results when matching, and unusable output in any other case, mimicking the mechanisms of publickey, private-key encryption. In doing so we are providing authors all the control associated with ICC profiles that are both device and media specific (in case of printers). The following section provides some background on ICC color management and specifically the elements that we employ for out method, the next two sections then detail our new approach showing results. Finally we conclude the paper by outlining the main benefits of the proposed method. Background Color management is a ubiquitous part of reproducing color content and at least for purposes of fine art and professional photography its employment can largely be taken for granted. Print service providers as well as authors themselves have been exposed for years to the inevitable necessity of controlled color in order to achieve repeatable, faithful and desired reproductions of photographic or fine-art color content. By far the most popular means to exercise control over color is that of the International Color Consortium’s (ICC) color management framework. Operating systems, color devices (cameras, printers, scanners), imaging software as well as some web browsers are now well equipped to handle content tagged with ICC profiles throughout. The ICC framework proposes the use of profiles associated with devices and/or content. It provides the ability to communicate color via a Profile Connection Space (PCS), representing colorimetry (e.g. CIE XYZ or LAB), the lingua franca among all proprietary device representations of color. Thus an image’s color is interpreted thanks to an associated source profile (e.g. sRGB) and employing a color management engine it can be transformed to a destination color space (e.g. some device CMYK) via the intermediate PCS. A fundamental principle of this workflow is that a device’s profiles are independent and agnostic of other devices and a transformation between any two is defined. [4] The key to this mechanism is thus the intermediate, common PCS. Figure 1. ICC Color Management communication via a common PCS. Each profile then provides a means to transform device color content into the PCS – the forward, AToB transform, and back into the device’s own color space – the reverse, BToA transforms. In order to encode these transformations, there are four mechanisms that can be employed: a 3x3 matrix (if the PCS is in CIE XYZ space), two linearization tables (or a gamma curve) that prefix and postfix a Color Look-Up Table Page 1 of 4
[1]
Brian A. Wandell,et al.
A spatial extension of CIELAB for digital color‐image reproduction
,
1997
.
[2]
Donald E. Knuth.
The art of computer programming: fundamental algorithms
,
1969
.
[3]
Donald E. Knuth,et al.
The Art of Computer Programming, Volume I: Fundamental Algorithms, 2nd Edition
,
1997
.
[4]
R. A. Fisher,et al.
Statistical Tables for Biological, Agricultural and Medical Research
,
1956
.
[5]
Ming Ronnier Luo,et al.
CIE 2000 color difference formula: CIEDE2000
,
2002,
Other Conferences.
[6]
J. Wishart.
Statistical tables
,
2018,
Global Education Monitoring Report.
[7]
M. Luo,et al.
The development of the CIE 2000 Colour Difference Formula
,
2001
.