Target Validation And Image Calibration In Scanning Systems
Recent Advances in Image, Audio and Signal ProcessingTarget Validation and Image Calibration inScanning SystemsCOSTIN-ANTON BOIANGIUDepartment of Computer Science and EngineeringUniversity “Politehnica” of BucharestSplaiul Independentei 313, Bucharest, 060042ROMANIAcostin.boiangiu@cs.pub.roALEXANDRU VICTOR ŞTEFĂNESCUComputer Science Department - MS 132Rice UniversityPO Box 1892, Houston TX 77251-1892USAstefanescu@rice.eduAbstract: - One of the directions for paper document conservation is conversion to microfilms and scannedimages. Since recently microfilming has been abandoned over digitization, there is a need for standards andguidelines for the conversion workflow. The article proposes a set of methodologies for calibrating scanningsystems to ensure high quality reproduction of both microfilms and original paper prints, in terms of tonalreproduction, geometric distortion and image sharpness.Keywords: Image calibration, sharpness, MTF, tonal reproduction, geometrical distortion, image acquisitionconverting pre-existing microfilms to digital media,in order to avoid handling the original documents, anaction which is both costly and potentially damagingfor the decaying prints.A further direction pursued in recent years isconverting scanned documents (either of themicrofilms or the original paper prints) intoelectronic files, especially for large electroniclibraries, for easier access to documents. Contentconversion systems, based on optical characterrecognition (OCR), enable operations such asediting, word searching, easy document storing andmultiplication, and the application of a large set oftext techniques including text-to-speech and textmining to be performed on the digitalized document.In addition, this ensures a better preservation oforiginal documents, due to minimizing the need forphysical use.Undoubtedly, digitization has many advantagesover microfilming on the access side: digital imagesinclude color reproduction, they allow remote accessand they can be easily searched by using OCR. Fromthe point of view of preservation, digitization offersexciting possibilities as well. Since digitized copiescontain color; a high quality digital image is closerto the original than a microfilm could ever be.1 Introduction and motivationPaper documents such as newspapers, books andother prints suffer in time of various forms ofautonomous decay that can affect paper, amongwhich are paper acidification and ink and coppercorrosion. In the 1980’s and 1990’s research wascarried out – e.g. the Metamorfoze project in theNetherlands, a collaborative effort of the KoninklijkeBibliotheek (National Library) and the NationaalArchief (National Archives) – to develop reliablemethods and standards for the conservation of paperheritage material that was considered of nationalimportance. The research focused on two directions:preservation – concerned with slowing down thedecay on the original documents through means ofdeacidification, treatment of ink corrosion andcopper corrosion, small repairs, acid-free wrappingsand climatized storage – and conversion – dealingwith the transfer of the threatened material to anotherstorage medium by means of either microfilming ordigitization.When research programsfirststarted,microfilming was a reliable method to preserve thecontent of an endangered document. However, inrecent years, digitization is preferred overmicrofilming, leading to an additional challenge ofISBN: 978-960-474-350-672
Recent Advances in Image, Audio and Signal ProcessingThere are nevertheless a number of issues to betackled before digitization can definitively be usedas a conversion method for preservation and access.Standards and guidelines, the workflow, themetadata, long-term storage and retrieval of digitalimages all have to be developed and dealt with.Scanning system calibration is one of the areasfor which accurate standards and methodologies areneeded. There are many factors which can affect thescanning quality: defective equipment, variations inillumination conditions, aging scanner lamps, out-offocus cameras, failing sensors on specific colorchannels, etc. This article proposes a set ofmethodologies for calibrating scanners to ensureoptimal quality in the digitization of both microfilmsand paper prints, covering the following issues: tonalreproduction and illumination, color cast and coloraccuracy, calibration and tonal reproduction, imagesharpness and optical distortion.2.1 Initial SetupThe monitor settings (e.g. white point 6500K,gamma 2.2, gray desktop background, etc.),workspace conditions (e.g. ambient illumination 3264 lux, color temperature 5000K, neutral ambientcolors, etc.) and scanning procedure should matchthe requirements mentioned in the Metamorfozeproject guidelines [5]. The aim of these standards isto remove any effects interfering with the subjective,visual assessment of the images, and to supportuniformity in assessment between supplier andclient.2.2 Target Sheet Composition and SequenceAll aspects are assessed by capturing images of aframe-filling white sheet of cardboard on top ofwhich various technical targets are placed. Theoptical density of the white cardboard must bebetween 0.05 and 0.15 [5]. For all target sheets, thedistance between them and the lens must be equal tothe distance between the original and the lens. Inother words: the reduction factor used for capturingthe target sheets much be equal to the reductionfactor used for capturing the originals.The following four target sheets are used insequence for evaluating document scanningperformance:First target sheet: tonal reproduction andillumination. Both aspects are assessed with the aidof one single image, which is constructed bycentering a Kodak Gray Scale (see Fig. 1) at thebottom of the cardboard sheet.2 Target ValidationScanner calibration is performed using specialtechnical targets. Target validation refers to theprocess of checking if certain parameters of thescanned images of the targets verify some predefinedstandards. There are two kinds of technical targets:targets that must be captured with every individualimage that is made from an original, and technicaltarget sheets that must be captured for every batch (aspecified number of images) or for a series of imagesmade in a specified period of time (for instance onemorning or afternoon) [5].Fig. 1. Kodak Gray ScaleSecond target sheet: color cast and coloraccuracy. The two aspects are evaluated using atarget sheet similar to the first, but with a color testISBN: 978-960-474-350-6target, the GretagMacbeth Color Checker SG (seeFig. 2), positioned in the center of the sheet.73
Recent Advances in Image, Audio and Signal ProcessingFig. 2. GretagMacbeth Color Checker SG: front, back and legendThird target sheet: sharpness. Again, the targetsheet is based on the first type, this time with fiveQA-62 slanted edge sharpness test targets (see Fig.ISBN: 978-960-474-350-63) placed in the center and the four corners of thetarget sheet.74
Recent Advances in Image, Audio and Signal ProcessingFig. 3. 5 x QA-62 targets plus Tiffen grayscaleFig. 4. QA-2 metric test targetFourth target sheet: optical distortion. The targetsheet is constructed by placing a QA-2 metric testtarget (see Fig. 4) containing length markers in thecenter of the white cardboard.For assessing microfilm scanning performance, aMicrofilm target sheet is used. It contains all testtargets from the four document scanning sheetsplaced on a single cardboard base. All aspects areISBN: 978-960-474-350-6evaluated by studying the scanned image of thetarget sheet captured on microfilm.For every individual image, it must be possible toassess tonal reproduction and color accuracy inrelation to the original. Therefore a Kodak GrayScale Q-13 or Q-14 and a mini GretagMacbeth ColorChecker Rendition Chart must be captured togetherwith every single original [5]. Both technical targets75
Recent Advances in Image, Audio and Signal Processingmust be positioned side by side, and clearly visible,centered at the bottom of the frame.How to enable assessment of color cast for eachindividual image is still being investigated [3]. Apossible solution might be to use a target with anumber of neutral gray patches, placed in a rightangle. This target could be positioned in each corner,and captured with each image.The following sections discuss the methodologyof evaluating each aspect in the scanning process.The proposed method is an improved version ofthe slanted edge method described in the ISO 12233methodology and the SAFECOM methods [4]. Theslanted edge method involves the analysis of a potionof an image containing an edge slightly tilted withrespect to the detector and, compared to othermethods, has the advantage of requiring a smallnumber of pixels from a single image to beprocessed.6. Sharpness Validation Methodology3. Tonal Reproduction andIlluminationThis section details a methodology suitable forcalibrating document or microfilm scanningequipment with respect to the image sharpnessquality factor.Measured on the basis of the Kodak Gray Scale (Q13or Q14) all patches of the Kodak Gray Scale shouldbe distinguishable from each other. The pixel valueof patch A has to be between 250-230. The pixelvalue of patch 19 must be above 10. The pixel valueshould be measured with a 5x5 average window. Fornoise test acceptance within the pixel values of theKodak Gray Scale (or equivalent) a maximumstandard deviation of 10 is allowed.6.1 Initial SetupFor measuring image sharpness, a special target shallbe constructed, containing five calibration targets(four near the corners and one in the center) such asthe QA-62 target (presented in Fig. 5) with fourslanted edges as sides of a rectangle. The target mustbe scanned and validated at regular intervals (e.g. atthe beginning of the day) or after any change inscanning parameters (e.g. resolution, scaling factorfor microfilms, etc.).4. Color Cast and Color AccuracyThe color cast is determined by measuring thegrayscale patches of the GretagMacbeth ColorChecker within a 5x5 average window. The patchesmust be neutral. The maximum deviation allowed is 4 or 4 pixel point difference between the RGBchannels for every patch, when taking the middleRGB-value as a starting point.5. Image Sharpness AssessmentThe sharpness of a photographic imaging system orof a component of the system (lens, film, imagesensor, scanner, enlarging lens, etc.) is a qualityfactor that determines the amount of detail that canbe reproduced. It is characterized by a parametercalled Modulation Transfer Function (MTF), alsoknown as spatial frequency response, which is ameasure of the response of an optical system tovarying intensities of light. The MTF is strictly theresponse to parallel lines whose brightness variesfrom minimum to maximum in a sinusoidal function.Traditional methods for MTF measurements wereinitially designed for devices forming continuousimages and can produce erroneous results, becausethe sampling of digital devices is not properly takeninto consideration [1]. Additionally, MTF results candepend on the chosen technique (sine target or bartarget utilization, slit or knife-edge technique).ISBN: 978-960-474-350-6Fig. 5. Quality Assurance 62 sharpness calibrationtarget6.2 Detecting the Region of InterestAutomated sharpness validation techniques can beapplied on the scanned target. To detect the fiveslanted rectangles in the target image, a conversionto black-and-white followed by 4-connected (black)pixel detection can be applied. By analyzing theshape of the connected regions, the rectangles canrecognized and their slant angles can be checked tomeet certain limits (e.g. between 2 and 5 degrees).For each of the five rectangles, image sharpnessshall be measured by processing the pixels contained76
Recent Advances in Image, Audio and Signal Processingin four regions of interest (RoI), corresponding to theslanted edges of the rectangle. The RoI is required tobe of a minimum size of 80 by 60 pixels (see Fig. 6)and is normally selected by dividing the minimalnon-slanted rectangle surrounding each slantedrectangle into 6 parts, both horizontally andvertically, and extracting 4/6 by 1/6 portions (e.g. forthe top edge, the RoI is situated in the top sixth andmiddle four sixths of the reference rectangle).horizontal Spatial Frequency Response (SFR) of thedetector.The ESF must be resampled to a fixed interval byaccumulating the projected pixels into “bins” havingthe width a fraction of the pixel pitch. This can beachieved by filtering the pixel values with a trianglefilter of unit height and the width of a bin. Thus, thevalue associated to each bin is the weighted averageof the pixels filtered by the triangle function centeredin the bin. This allows analysis of spatial frequenciesbeyond the normal Nyquist frequency [4]. Thenumber of bins per pixel distance is usually chosenas 4. Higher values may lead to insufficientlypopulated or empty bins.Fig. 6. Best minimum cropped region of nFig. 7. (a) ESF, (b) LSF, (c) Hamming LSF, (d) MTFplotsThe algorithm for computing the MTF and theassociated frequency response graph is derived fromthe International Standard 12233 [4]. The followingsteps are performed for each RoI of each QA-62target and, depending on the employed scanningcolor space, for each RGB color channel pluscombined luminance channel (Y 0.299 Red 0.587 Green 0.114 Blue) for document scans,or just the gray channel for grayscale microfilmscans:For each pixel column in the RoI (which isrotated to the position corresponding to the top edgeRoI, for reference purposes) the position of theseparation line between the background and theslanted rectangle is determined by maximizing thedifference of the sum of weighted pixel values on thetwo sides of a triangle filter of predefined width (e.g.10 pixels) sliding over the pixels in the column.The least-squares fit line through the coordinatesfound at point (1) is determined and is used toapproximate the separation border between thebackground and the slanted rectangle.Pixels in the RoI no further than a predetermineddistance (normally 1mm, around 12 pixels at300DPI) from the fitted line, on both sides, areprojected along the edge transition, resulting indistance-color tuples. These values represent theEdge Spread Function (ESF) which is the systemresponse to the input of an ideal edge [2]. The ESF issuper-sampled because of the slanted edge whichinduces differences in the sub-pixel location of theprojected pixels onto the perpendicular. A verticallyoriented edge would only allow obtaining theISBN: 978-960-474-350-6The equally spaced ESF samples obtained at (4)are derived (d/dx) in order to obtain the Line SpreadFunction (LSF). A Hamming windowing function isapplied to force the derivative to zero at theendpoints [4], reducing the effects of the Gibbsphenomenon that results from truncation of aninfinite series [1].The normalized magnitude of a linear FastFourier Transform performed on the LSF yields theMTF (see Fig. 7).Care must be taken in selecting the number ofpoints calculated along the ESF with respect to thesampling rate in order to obtain the desired numberof points in the resulting MTF. The frequency axis ofthe MTF must be scaled to represent the calculatedMTF in terms of the Nyquist frequency of theimaging system, defined as the highest sinusoidalfrequency that can be represented by a sampledsignal and is equal to one half the sampling rate ofthe system [2] – always 0.5 cycles per pixel.Formaximumprecisioninsharpnessmeasurement, steps (3) to (6) in the MTFcomputation algorithm can be repeated for theinterpolated line at step (2) rotated by slight angles insteps of 0.1 degrees, taking into consideration onlythe MTF curve with the highest values.6.4 Sharpness SpecificationFor a scanning system to pass sharpness validationcertain criteria must be defined. Relevant indications77
Recent Advances in Image, Audio and Signal Processingare found by checking the frequency at which theMTF graph drops to 10% of its initial, zerofrequency value. Values above 70% of the ding to half the maximum MTF value(MTF50P) is also a good sharpness metric.Furthermore, internal sharpening (performed byfirmware in scanning equipment) can be detected bycomparing the peak MTF value with the initial value.A ratio below 1.2 is acceptable [5].Modulation Transfer Function of the scanningsystem.References:[1] [1] Magali Estribeau ; Pierre Magnan, FastMTF measurement of CMOS imagers usingISO 12233 slanted-edge methodology,Proceedings of SPIE, Vol 5251, 2004[2] [2] Kohm, K. Modulation Transfer FunctionMeasurement Method and Results For theOrbview-3 High Resolution ImagingSatellite, Proceedings of ISPRS XXXV,Istanbul, July 2004, pp. 7-12[3] [3] McCamy, C.S., Marcus, H., andDavidson, J.G., A Color Rendition Chart,JournalofAppliedPhotographicEngineering 11(3) (Summer issue, 1976), pp.95-99.[4] [4] SAFECOM. Public Safety Statement ofRequirements for Communications &Interoperability, The SAFECOM ProgramDepartment of Homeland Security, pg.99103, Vol II, version 1.0, August 2006[5] [5] Metamorfoze. Metamorfoze ek: National Library of theNetherlands, The Hague, June 20077. Optical DistortionsThe allowed deviation is a change in length or heightof 1% at the most. The Image Evaluation Test Target(QA-2) must be used. The size of this target is A3.To measure larger sizes, a larger test target must beused.8. ConclusionsDigitalization is the future for the preservation ofinformation contained in decaying paper prints.Detailed methodologies for calibration of scanningequipment are required to avoid geometric and colordistortions, as well as ensuring a level of high imagesharpness. The paper presented a methodology andan algorithm to assess sharpness based on theISBN: 978-960-474-350-678
centering a Kodak Gray Scale (see Fig. 1) at the bottom of the cardboard sheet. Fig. 1. Kodak Gray Scale . Second target sheet: color cast and color accuracy. The two aspects are evaluated using a target sheet similar to the first, but with a color test target, the GretagMacbeth Color Checker SG (see
Fig. 1. Illustration of placing the calibration target at different dis-tances from the camera. Compared to putting the calibration target at the focal plane (Z 2), the calibration target dimensions could be sub-stantially smaller if the calibration target is placed at the out-of-focus plane (Z 1). However, if the target is placed at Z 1, the .
Calibration (from VIM3) Continued NOTE 1 A calibration may be expressed by a statement, calibration function, calibration diagram, calibration curve, or calibration table. In some cases, it may consist of an additive or multiplicative correction of the indication with associated measurement uncertainty. NOTE 2 Calibration should not be .
The target validation process is composed of several steps, from target reception to the communication of final decisions and feedback. The target validation process falls under the SBTi target validation service. Under this service, there are two distinct validation options available, that depend on the size of the company:
Host action Target needed * : Timings are given for information only, they can vary depending on the Host capabilities Calibration Data result Calibration data Calibration data Calibration data Calibration data Device initialization SPADs calibration Temperature calibration Offset calibrat
1 Manual calibration introduction 3 1.1 Purpose 3 1.2 Manual calibration software 3 2 Manual calibration overview 3 3 Manual calibration process 4 3.1 Setup MCR program 4 3.2 Calibration pattern positioning 9 3.3 Capturing pattern image 14 3.4 Using the Synergy Calibration tool 18 3.5 Fine tuning side images 24
Therefore, target 1 has three target drops, i.e., target 1-A-1, target 1-B-1 and target 1-C-2. In this manner we can enumerate all possible target drops from target information. From source and target information we can set all possible assignments, and each of them is composed of a source and sequence of target drops, called a target drop set .
calibration. The vertical comparator was built during the year 2003. The calibration facility is designed to calibrate up to 3-m-long invar rods, both for system calibration of digital levels and for tradi-tional rod calibration. SLAC System Calibration Facility The procedure of system calibration of digital levels is described
Calibration C-0942, Mechanical(Dimensional, Pressure, Mass & Volume) Calibration C-0943 & Thermal Calibration C-0944. Our company is also ISO 17025:2005 Certified for Mechanical Calibration, Elctro Technical Calibration & Thermal Calibration. The firm is proud to have earned distinction in its various services and we are performing as a
