1、 ISO 2012 Graphic technology and photography Colour characterization of digital still cameras (DSCs) Part 2: Considerations for determining scene analysis transforms Technologie graphique et photographie Caractrisation de la couleur des appareils photonumriques Partie 2: Considrations pour dterminer
2、 les transformations danalyse de scne TECHNICAL REPORT ISO/TR 17321-2 First edition 2012-10-15 Reference number ISO/TR 17321-2:2012(E) ISO/TR 17321-2:2012(E)ii ISO 2012 All rights reserved COPYRIGHT PROTECTED DOCUMENT ISO 2012 All rights reserved. Unless otherwise specified, no part of this publicat
3、ion may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or ISOs member body in the country of the requester. ISO copyright office Case postale 56 CH-1211 Geneva 2
4、0 Tel. + 41 22 749 01 11 Fax + 41 22 749 09 47 E-mail copyrightiso.org Web www.iso.org Published in Switzerland ISO/TR 17321-2:2012(E) ISO 2012 All rights reserved iii Contents Page Foreword iv Introduction v 1 Scope . 1 2 Definitions . 1 3 Goals 4 4 Fundamental colour-related DSC characteristics 5
5、4.1 Camera gain . 5 4.2 Camera dark current 5 4.3 Focal plane opto-electronic conversion function (FP OECF) 6 4.4 Camera opto-electronic conversion function (Camera OECF) . 6 4.5 Camera flare average percent. 6 4.6 Camera spectral sensitivities (including non-removable optical elements) . 6 4.7 Remo
6、vable optical element spectral transmittances . 6 5 Scene analysis transform parameters 7 5.1 Determined using spectral measurements . 7 5.2 Determined using test targets . 8 6 Scene analysis transform determination using spectral measurements 9 7 Scene analysis transform determination using test ta
7、rgets .10 7.1 Procedure .10 7.2 Test target characteristics 12 7.3 Capturing images of test targets .12 8 Applying the scene analysis transform and encoding 13 9 Considerations for selecting scene analysis transforms 14 10 Suggested metadata for scene-referred colour encodings 14 10.1 Camera charact
8、eristics 14 10.2 Colour encoding characteristics 14 10.3 Image specific characteristics .14 Annex A (informative) Example calculation results 15 Bibliography .17 ISO/TR 17321-2:2012(E) Foreword ISO (the International Organization for Standardization) is a worldwide federation of national standards b
9、odies (ISO member bodies). The work of preparing International Standards is normally carried out through ISO technical committees. Each member body interested in a subject for which a technical committee has been established has the right to be represented on that committee. International organizati
10、ons, governmental and non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization. International Standards are drafted in accordance with the rules given in the
11、 ISO/IEC Directives, Part 2. The main task of technical committees is to prepare International Standards. Draft International Standards adopted by the technical committees are circulated to the member bodies for voting. Publication as an International Standard requires approval by at least 75 % of t
12、he member bodies casting a vote. In exceptional circumstances, when a technical committee has collected data of a different kind from that which is normally published as an International Standard (“state of the art”, for example), it may decide by a simple majority vote of its participating members
13、to publish a Technical Report. A Technical Report is entirely informative in nature and does not have to be reviewed until the data it provides are considered to be no longer valid or useful. Attention is drawn to the possibility that some of the elements of this document may be the subject of paten
14、t rights. ISO shall not be held responsible for identifying any or all such patent rights. ISO/TR 17321-2 was prepared by Technical Committee ISO/TC 42, Photography. ISO 17321 consists of the following parts, under the general title Graphic technology and photography Colour characterization of digit
15、al still cameras (DSCs): Part 1: Stimuli, metrology and test procedures Part 2: Considerations for determining scene analysis transformsiv ISO 2012 All rights reserved ISO/TR 17321-2:2012(E) Introduction Digital still cameras (DSCs) have become the predominant means of photographic image capture, bu
16、t the nature of the image data produced by different cameras, or even by the same camera operating in different modes, is quite variable. This variability can cause problems in workflows, miscommunications and interoperability issues. This Technical Report provides information about methods for dete
17、rmining scene analysis transforms, which are transforms that convert raw image data to scene-referred image data. This information is provided in the form of a Technical Report because there are a number of choices to be made when determining scene analysis transforms. These choices are influenced b
18、y the subject matter being photographed (including the scene illumination), the scene adopted white and the adopted white of the scene-referred colour encoding to be used, aesthetic choices regarding scene analysis colour error minimization, and other considerations. It is not possible to provide mo
19、re specific recommendations because the spectral responses of DSC colour analysis channels do not, in general, match those of a typical human observer, such as defined by a CIE standard colourimetric observer. Nor do the responses of different DSCs ordinarily match each other. This Technical Report
20、outlines considerations relevant to the determination of scene analysis transforms based on the minimization of errors in specified colour spaces. The DSC characterization data obtained using ISO 17321-1 serve as the raw DSC image data. Good understanding of this Technical Report requires that the t
21、hree fundamental modes of DSC operation be distinguished: the raw mode, the scene-referred mode, and the output-referred mode. When operating in the raw mode, a DSC records image data that is most closely related to the sensor response. Some types of processing may have been performed, such as dark
22、current subtraction, defect removal and colour filter array interpolation, but neither a scene analysis nor a colour rendering transform has been applied. Any encoding transform typically consists of only a non-linearity to better align the quantization intervals with the image noise characteristics
23、, and possibly some form of compression. When operating in the scene-referred mode, a DSC records image data that represents an estimate of the scene or focal plane image relative colourimetry, typically with white balancing to the encoding adopted white. The image data has not undergone colour rend
24、ering for some anticipated output medium and viewing conditions. In order to produce output-referred images intended for reproduction, it is necessary to either apply a colour rendering transform directly to the scene-referred images, or convert them to a working colour space where the desired colou
25、r rendering is applied. Camera controls or raw processing software can offer some aesthetic choices when converting to scene-referred, but the results of such choices need to be viewed through the intended colour rendering transform in order to see their effect on the final output. The image data ar
26、e encoded prior to applying the colour rendering transform, and are therefore not an encoding of the intended output colourimetry. At present, few DSCs offer an in-camera scene-referred mode, although some camera raw processing applications have this capability. When operating in the output-referred
27、 mode, the DSC controls are set to achieve the desired output directly, thereby incorporating the colour rendering, and in many cases a reference output device encoding (such as for a CRT monitor) in the image file. When operating in this mode the DSC encodes the colourimetry of the intended output
28、on the reference medium, not scene-referred colourimetry. Also, the output-referred colourimetry can be in different encodings, with different reference media, and in some cases will need to be colour re-rendered and/or re-encoded to produce different reproductions. The information provided in this
29、Technical Report is intended to help camera and raw processing software manufacturers, professional photographers and colour measurement applications to determine, communicate about, and select DSC scene analysis transforms. However, it will often not be practical for end users to determine scene an
30、alysis transforms themselves. In addition to the requirement for raw DSC image data, relatively sophisticated and expensive measurement equipment is required to obtain chart patch spectral reflectance or radiance, illumination source spectral power, and DSC spectral sensitivity (as described in ISO
31、17321-1). ISO 2012 All rights reserved v ISO/TR 17321-2:2012(E) Exposing Light A/D Converter Image representation in raw digital camera response coordinates Transform Image representation in scene-referred colour encoding Colour rendering Image representation in output-referred colour encoding Trans
32、form Image representation in printer device coordinates Sensor device dependent device dependent device independent device independent Fig u r e 1 G en er i c imag e wor k f lo w f or d i g it a l phot o g r aph yvi ISO 2012 All rights reserved TECHNICAL REPORT ISO/TR 17321-2:2012(E) Graphic technol
33、ogy and photography Colour characterization of digital still cameras (DSCs) Part 2: Considerations for determining scene analysis transforms 1 Scope This Technical Report provides information about methods for determining scene analysis transforms based on the minimization of errors in estimated sce
34、ne or focal-plane colourimetry, including corresponding colourimetry. These transforms are limited in applicability to raw DSC image data. This Technical Report concerns only the creation and encoding of scene-referred and focal-plane-referred image data. It does not address the encoding of output-r
35、eferred image data. It also does not provide information relating to the specification of metadata items describing intended artistic adjustments, colour rendering and viewing. This Technical Report does not address how to choose adopted white points or how to process scene- referred image data to p
36、roduce output-referred image data. This Technical Report is not intended to be comprehensive or complete; it is an overview intended to enable improved practices and communications. 2 De finiti ons For the purposes of this document, the following definitions apply. 2.1 absolute colorimetric coordina
37、tes tristimulus values, or other colorimetric coordinates derived from tristimulus values, where the numerical values correspond to the magnitude of the physical stimulus EXAMPLE If the colourimetric coordinates used are CIE 1931 standard 2 observer tristimulus values, the Y value should correspond
38、to the luminance, not the luminance factor (or some scaled value thereof). 2.2 adapted white colour stimulus that an observer who is adapted to the viewing environment would judge to be perfectly achromatic and to have a luminance factor of unity; i.e. absolute colorimetric coordinates that an obser
39、ver would consider to be a perfect white diffuser NOTE 1 The adapted white can vary within a scene. NOTE 2 No assumptions should be made concerning the relation between the adapted white and measurements of near perfectly reflecting diffusers in a scene, because measurements of such diffusers will d
40、epend on the illumination and viewing geometry, and other elements in the scene that can affect perception. It is easy to arrange conditions for which a near perfectly reflecting diffuser will appear to be grey or coloured. NOTE 3 See adopted white (2.3). ISO 2012 All rights reserved 1 ISO/TR 17321-
41、2:2012(E) 2.3 adopted white spectral radiance distribution as seen by an image capture or measurement device and converted to colour signals that are considered to be perfectly achromatic and to have an observer adaptive luminance factor of unity; i.e. colour signals that are considered to correspon
42、d to a perfect white diffuser NOTE 1 The adopted white can vary within a scene, if such variation is supported by the imaging system. NOTE 2 The adopted white is not required to be an estimate or approximation of the adapted white. For example, if a scene lit by tungsten illumination is captured usi
43、ng a DSC with the white balance set to D55 (daylight), the adopted white will be D55 but the adapted white will be closer to a tungsten illuminant (e.g. ISO 7589 Studio Tungsten or CIE Illuminant A). NOTE 3 See adapted white (2.2). 2.4 scene analysis transform spectral limit mapping of spectral (mon
44、ochromatic) colours as captured by a DSC and transformed to a scene-referred colour encoding using a specified scene analysis transform 2.5 colour component transfer function CCTF single variable, monotonic mathematical function applied individually to one or more colour channels of a colour space N
45、OTE 1 Colour component transfer functions are frequently used to account for the non-linear response of a reference device and/or to improve the visual uniformity of a colour space. NOTE 2 Generally, colour component transfer functions will be non-linear functions such as a power-law (i.e. “gamma”)
46、function or a logarithmic function. However, in some cases a linear colour component transfer function will be used. 2.6 colour gamut solid in a colour space, consisting of all those colours that are either: 1) present in a specific scene, artwork, photograph, photomechanical or other reproduction;
47、2) capable of being created using a particular output device and/or medium NOTE See scene analysis transform spectral limit (2.4). 2.7 colour matching functions tristimulus values of monochromatic stimuli of equal radiant power CIE Publication 17.4, 845-03-23 2.8 colour pixel reconstruction algorith
48、m that creates a fully populated colour image record from the output of a colour filter array type sensor by interpolating values for each colour at each pixel location, also known as demosaicing or colour pixel interpolation 2.9 colour rendering mapping of image data representing the colorimetric c
49、oordinates of the elements of a scene or original to image data representing the colorimetric coordinates of the elements of a reproduction NOTE Colour rendering generally consists of one or more of the following: compensating for differences in the input and output viewing conditions, tone scale and gamut mapping to map the scene colours onto the dynamic range and colour gamut of the reproduction, and applying preference adjustments.2 ISO 2012 All rights reserved ISO/TR 17321-2:2012(E) 2.10 colour space geometric representation of
