SMPTE RP 133-1991 Specifications for Medical Diagnostic Imaging Test Pattern for Television Monitors and Hard-Copy Recording Cameras《电视监视器和截屏图录制摄像机医疗诊断成像试验模式规范》.pdf

1、1 ScopeThis practice describes the format, dimensions, andcontrast required to make diagnostically significantmeasurements of the display and camera systemresolution for both digital and analog monochromesignal sources. (See note.) The practice providesusers of medical diagnostic imaging systems wit

2、h acomprehensive test pattern for day-to-day operationalchecks and adjustments of focus, brightness andcontrast, resolution response, mid-band streaking,uniformity, and linearity of viewing monitors and hard-copy recordings. This practice is not intended to cre-ate a standard for image characteristi

3、cs such asresolution, geometry and linearity, uniformity, phos-phor defects, etc. However, use of the pattern isencouraged as an appropriate tool for evaluating themeasurement and specification of such image char-acteristics.2 Description2.1 PatternA reproduction of the test pattern is shown in figu

4、re 1.Figure 2 is a drawing of the pattern for reference tothe following text:2.2 BackgroundThe background is presented at 50% of maximumwhite to obtain an approximate average picture level(APL) of 50% (figure 2 1).2.3 Crosshatch patternThis pattern allows verification of linearity, either byusing a

5、straightedge or by overlaying or projecting aperfect reference image (Ball chart) (figure 2 2).2.4 Resolution patternThree zones of vertical and horizontal, high-contrastbar patterns are placed in five locations (four cornersand the center). The modulation depth of the patternis 100%. Placement of t

6、he patterns facilitates com-parison of the resolution in the center and in thecorners of the screen (figure 2 3). A detailed view ofthe resolution pattern is shown in figure 3.2.5 Low-contrast imaging resolutionThree zones of vertical and horizontal, low-contrast,equally-spaced bars are located next

7、 to the patternsdescribed in 2.4. The modulation depth of these areasis 1%, 3%, and 5%. This is the most sensitive test forvarious types of imaging noise (figure 2 4).2.6 Gray step patternBlocks of specified signal amplitude, representing 0%to 100% of the input signal in steps of 10%, allowmeasureme

8、nts of the response characteristics of theviewing monitors and hard copy (figure 2 5).2.6.1 Small-contrast changesSmall-contrast changes of 5% of the input signal aresuperimposed on a 0% background, and a 95% valueis superimposed on a 100% maximum signal back-ground to allow quick verification that

9、all availablegray-scale information can be presented in the imageto be displayed (figure 2 6).2.7 White window and black windowDuring display of these patterns, the display or record-ing device signal stays relatively long in one state(black or white). These large areas of maximum con-trast facilita

10、te detection of mid-band streaking (poorlow-frequency response), video amplifier ringing orovershoot, deflection interference, and halo (figure 27 and 8).Specifications for Medical Diagnostic Imaging Test Pattern forTelevision Monitors andHard-Copy Recording CamerasRevision of RP 133-1986RP 133-1991

11、SMPTE RECOMMENDED PRACTICEPage 1 of 6 pagesApproved March 1, 1991Copyright 1991 by theSOCIETY OF MOTION PICTURE AND TELEVISION ENGINEERS595 W. Hartsdale Ave., White Plains, NY 10607(914) 761-1100REAFFIRMED 1999 Figure 1 Video hard-copy camera image of test patternFigure 2 Schematic drawing of test p

12、attern in a square formatRP 133-1991Page 2 of 6 pages2.8 Picture borderA line is placed around the outside of the pattern thatdefines the full extent of the pattern (figure 2 9).3 Format3.1 The test pattern (figure 2) and the followingparagraphs shall represent a recommended im-plementation of the p

13、attern. Specific informationfor digital implementation is given in table 1.3.2 The middle gray background is used to iso-late the gray-scale and center resolution patternfrom interference by adjacent patterns.3.3 Crosshatch lines in the horizontal directionshall be two raster lines wide, and in the

14、verticaldirection, lines shall be of a width equal to tworaster lines. The intensity of the line shall be75% of the maximum signal. The patterns shownconsist of square spaces equal to 10% of theraster height. Rectangular aspect ratios and dif-ferent numbers of scan lines may require frac-tional spac

15、es (see figure 4). Fractional spaces,if required, should be at the edges of the pattern.3.4 The high-contrast resolution patterns oflines in the vertical direction are divided intothree zones of different spatial frequencies withthe highest frequency pattern in the four corners.The center pattern is

16、 oriented with its high-con-trast, high-resolution pattern in the upper-rightcorner (see figure 3). Modulation of the high-contrast pattern is 100%.3.5 It is recommended that the high-contrastmiddle spatial frequency zone of the horizontalresolution patterns (which are placed in the fourcorners of t

17、he test pattern) should represent thecorner resolution specified by the manufacturerfor the system being evaluated. The high-con-trast middle frequency zone of the resolutionpattern (in the center of the test pattern) shouldrepresent the central display resolution as speci-fied by the manufacturer o

18、f the system beingevaluated. The lower frequency pattern in eachof these five high-contrast test patterns shouldbe 75% of the resolution specified by the manu-facturer for the area in which the resolution pat-tern is located. The high-contrast highestfrequency pattern should be 125% of the resolu-ti

19、on specified by the manufacturer for the areain which the pattern is located.3.5.1 Vertical resolution patterns are limited tothree arrangements; one line off and one line on,two lines off and two lines on, and three lines offand three lines on.3.6 For digital systems, the spatial frequency ofthe ba

20、r patterns is limited to increments of onepixel size. Thus, it is recommended that onepixel on and one pixel off be used for the highestresolution zone and three pixels on and threepixels off be used for the lowest resolution zone.3.7 Low-contrast patterns repeat the format ofthe high-contrast patte

21、rns except the spatial fre-quency in all three zones is equal to that of theabove middle zone; however, in one zone themodulation is 1% (i.e., white equals 51% andblack equals 50%). In the middle zone, themodulation is 3% (51% and 48%), while in thelast zone, it is 5% (53% and 48%). The arrange-ment

22、 of patterns is symmetrical, with the lowest-contrast zones closest to the corners.3.8 A rectangular area 40% of the picture heightis used to display at least 10 steps of luminancefrom white to black. Thus, the resolution per-formance is evaluated at the proper dynamicrange of the image as it is use

23、d for operation forboth viewing monitor and film hard copy. Theindividual blocks are intended to be measuredwith a spot photometer or optical densitometer,and they are concentrated in the middle of thescreen to avoid shading caused by off-axis ef-fects in the CRT or optical system.3.9 Small contrast

24、 changes in the blocks next tothe white and black ends of the gray scale con-sist of a 5% maximum signal patch superim-posed on 0% (i.e., black) and a 95% maximumsignal patch superimposed on 100% (i.e., white)to permit fine tuning of contrast and brightnesscontrols to verify that all shades of gray

25、presentin the signal can be perceived in the image. Thesize of the inset shall be 5% of the picture height.3.10 The white and black window patternsshould have a height equal to 8% of the pictureheight centered in one square of the crosshatchwith the length starting one square away fromRP 133-1991Pag

26、e 3 of 6 pagesFigure 3 Detail of digitally-generated resolution pattern portion oftest pattern showing high- and low-contrast portionsFigure 4 Schematic drawing of test pattern in rectangular formatRP 133-1991Page 4 of 6 pagesArray sizePercentvideosignalPercent ofpictureheight256x256320x320512x51210

27、24x10242. Crosshatch vertical spacing 75% 10% 25 32 50 102horizontal spacing 10% 25 32 50 102line width 2 2 2 23. High-contrast resolution Each block 6.25% 16 20 32 64Low frequency 3 pixels on,3 pixels offMid frequency 2 pixels on,2 pixels offHigh frequency 1 pixel on,1 pixel offContrast all the sam

28、e 0-100%4. Low-contrast resolution Each block 6.25% 16 20 32 64Frequency 2 pixels on,2 pixels offContrast low 50-51%medium 48-51%high 48-53%5. Gray scale pattern 10% 25 32 50 1026. Small contrast changes 5% and 95% 5% 12 16 25 517. however, it can be utilizedequally well for color displays and monitors in the RGB orencoded mode of operation.RP 133-1991Page 6 of 6 pages