Presentation on theme: "HIGH-PRECISION MEASURING SCALE RULERS FOR SCANNERS HIGH-PRECISION MEASURING SCALE RULERS FOR SCANNERS E.V. Poliakow 1, V.V. Poliakov 1, L.A. Fedotova 2,"— Presentation transcript:
HIGH-PRECISION MEASURING SCALE RULERS FOR SCANNERS HIGH-PRECISION MEASURING SCALE RULERS FOR SCANNERS E.V. Poliakow 1, V.V. Poliakov 1, L.A. Fedotova 2, M.K. Tsvetkov 3 (1)Central (Pulkovo) Astronomical Observatory (GAO RAN) ; (2) Leningrad Optical-Mechanical Association (LOMO) (3) Institute of Astronomy, Bulgarian Academy of Sciences, Bulgaria E.V. Poliakow 1, V.V. Poliakov 1, L.A. Fedotova 2, M.K. Tsvetkov 3 (1)Central (Pulkovo) Astronomical Observatory (GAO RAN) ; (2) Leningrad Optical-Mechanical Association (LOMO) (3) Institute of Astronomy, Bulgarian Academy of Sciences, Bulgaria ABSTRACT: Digitisation of astrographic plates from the astrometric series using the flatbed scanners and microdensitometers does not preserve high metrological characteristics of the read-out information due to construction features of these devices. The application of the proposed high-precision measuring scale rulers allow to transform a scanner into an efficient measuring device not conceding of their precision characteristics similar to the measuring engines of the ASCORECORD type. Conclusions: Conclusions: In the present paper was disused the problems raised during performed and planned measurements of the electronic copies i.e digitization of the astronomical plate collections stored at the Russian astronomical observatories. As the problem is pretty actual we decided briefly to summarize our preliminary results of using scale-rulers, before the detailed investigations of the proposed method. We propose to use a flatbed scanners completed additionally with especially calibrated glass-rulers in order to improve the results of more precise astronomical plate measurements. The preliminary tests shows that using this method we increase the accuracy of the plate digitization. Fig. 1: The Pulkovo UMAX 1200 flatbed scanner (a); the lamp of the transparency mode is ”ON” (b). Fig. 2. The positioning systems of the transparency mode lamp (a) and a CCD (line) imager as a light-detector (b). REFERENCES  Resolution B3 of the 24th IAU GA, Manchester (2001) The IAU Inform. Bull., 88, 40.  Pratt N. M. (1970), in: Proceedings of the IAU Colloquium No. 11, Edinburgh,  Pier J. R. and Monet D. G. (1993) in: Workshop on Databases of Galactic Structure, eds. A.G.D Philip, B. Hauck, A. I. Upgren, L. Davis Press,  Pratt N. M. (1977) Vistas in Astronomy, Pergamon Press., 21,  Поляков Е. В., Герасимов А. Г., Пикин Ю. Д., Савастеня А. В., Соколов A.B (1994), Измерительная техника, № 3, М.,  Поляков, Е. В., Канаева, Н. Г., Канаев, И. И., Пугач, Т. Н. (2002) Изв. ГАО РАН, 216, СПб,  Измайлов И. С. (2000) Изв. ГАО РАН, 214, СПб,  Barbieri C., et al. (2003) Experimental Astronomy, 15, Fig. 9. A scheme of the bench-marking of a standard scale for a scanner Fig. 7: The plastic particles of the worn-out axle of the tension pulley sticking to the electrified glass of the adapter Fig. 6: The plastic particles of the worn-out adapter supports which have got into inner space of the scanner a b Пулково-ГАО, РАН b a SCALE'S SERTIFICATION AND APPLICATION : The manufactured glass-scales (Fig. 9.) are being certified by means of highly accurate measurements of positions of slant lines and the solid scale line (the guide) using AMC FANTAZIA. The electronic certificate is done for each scale, being a file containing the corrections of the guide for non-rectilinearity, the line coordinates relative to the guide, the coordinates of singular points, i.e. hindrances, flaws, increased thickness and gaps in marks and lines. The scale certificates are used by the program for correcting the digitized image.The scales are laid parallel with the plate edges by sufficiently free tolerance. The difference between the scale zero-points and their deviation from the Y-axis direction are determined and compensated by a program while processing. An plate is scanned together with the scales, constituting the unified digital image with them. The image correction is based on the scale measurements and done whether directly by the position measurements of objects or on the stage of preparation of an image for measurements, its corrected version being obtained. There are “pro” and “contra” arguments for each approach. In the first case none image element is changed, the file contents being untouched, every point constituting the object under measurement takes part in the measurement process having its own density value and precise coordinates as well. As to the second approach, a new file is created, the elements being the results of the bi-linear or bi-cubic interpolation which leads to some smoothing of images imperceptible practically on objects which are the characteristic ones for astro-plates. This processing version is efficient when extracting individual images from the common picture or when several images are commonly processed because the corrections are introduced into the coordinates of everyone of them. INTRODUCTION In the recent time the astronomical world community is concerned with the problem of preservations the photographic observation materials having almost one and a half century history . The automated measuring complexes (the AMC) such as PDS, GALAXY, PMM, COSMOS, FANTAZIA, etc.was designed in early 1980 to solve this problem [2-5]. These systems, however, run to units in the whole world and are beyond the reach for most of the astronomical plate collections due to the fact that the AMC are stationary installations, and to convey great amounts of tens of thousands of astro-plates weighing tons to those installations to be processed turns as a rule to be an unrealizable undertaking. There is only one high-class plate measuring device in Russia - the AMC “FANTAZIA” installed at the Pulkovo Observatory. This precise machine was dedicated to digitize the observatory almost 50,000 plates - which is about 25-30% of all astronomical photographic observation materials accumulated in Russia. In the first stage of the project aiming at digitalization of the total contents of the Observatory’s astro-plate collection and at saving it on the electronic data media has been carried out at the Laboratory of scientific measurements of the Observatory . The flat-bed scanners of UMAX-1200 and UMAX-2400 types with an adapter for transparent data media were used in this work (Figs. 1, 2). The scanning operation has been performed with the resolution of 600 through 1200 dpi. The electronic copies of astro-plates are the preparatory material for the second stage, i.e. for the further, high-precision digitalization of the contents of the astro-plates collection to be carried out by use of the AMC FANTAZIA. Several years was needed to complete this project. Similar work has been initiated since 2004 for the largest Russian astronomical photographic archives at the Institute of Astronomy of the RAS (INASAN), the Sternberg Astronomical Institute of the Moscow State University. The digitalization is planned to be carried out by use of high-resolution flatbed scanners as EPSON 1640XL and CREO. That way we need a high resolution accuracy of scanner and repeatability, however, the known flatbed scanner still are not guarantee the needed accuracy to preserve and reach the characteristics of the photographic material. It’s mainly due to the construction features of this scanners, i.e. because of lack of the light detector (the CCD ruler) positioning check during the scanning process. CONSTRUCTION FEATURES OF THE FLATBED SCANNERS The scanner for the digitization of transparent materials (f.e. Fig. 1) is composed of two functionally similar blocs, i.e. of the scanner itself and the adapter for transparent plates plates/films measurement. The systems for the frame illuminator positioning (the adapter, Fig. 2a) and this of the light detector (CCD imager) (the scanner, Fig.2b) are mounted in the both blocs and the movement of which during the scanning process is synchronized. The light source contains a gas-discharge lamp of high luminosity, and the light detector consists the mirror-lens system, slit diaphragms, and a CCD-line imager. Both carriages are moved by step-motors in the faltering mode, each of them being shifted parallel to itself. The parallelism is secured by bushings which are impressed into the carriages and sliding along the steel staves. The positions of the staves of the illuminator and of the light detector is different, i.e. they are close to the middle of the carriage, and to one edge of the the another one. The other edge being free and moving on the horizontal shelf leaning on a roller. The application points of the traction are situated near to the bushings. As a transmitter the cogged rubber belt is used. The carriage position is defined by a number of the motor step. The scans are formed at the pause instants in the movement. The scanning step along the X-axis is defined, therefore, by the optical characteristics of the device, whereas that along the Y-axis by the mechanical ones. The working field of the scanner and of the adapter are confined by glass plate; the volumes of both blocs being protected considerably from the soiling from outside. Fig. 8. A real bug inside the scanner THE SOURCES OF ERRORS DURING THE SCANNING The X co-ordinate errors: The non-rectilinearity, or the flexure of the guide in the vertical plane, |z| < 30 mm, the CCD-line imager would consequently be inclined with respect to a scanned astro-plate plane by the angle y ≈ Δz / L, y << 1°, the error Δx ≈ hy, L is the length of the CCD-ruler, h ≈ 45 mm is the distance of the astro-plate from the prism. The non-rectilinearity of the guide in the horizontal plane by the Δx value which leads consequently to the shift of the ruler by Δx, |Δx|max < 10 ≈ 15 mm. The eccentricity of the driven roller on the loose end of the light-detector carriage, the effect being the same as above. Thus are main sources of the errors, without optical ones (now not taken in the account which can performed the image distortion along the X-coordinate). The image scan-out along X axes being performed by the CCD-ruler, the shift of a line at all. without change of mutual positions of its individual elements which can to be considered as a characteristic error. On the contrary, the mechanical scan-out is generating the Y-distortions connected with the skews of the carriage while moving and with the positioning errors. The Y co-ordinate errors: The non-rectilinearity of a guide in the horizontal plane by a value of Δxb, b being the distance between driving bushings (see Fig. 2b). The effect is the ruler skew by the angle j ≈ Δxb / b, j << 1°, |Δxb|max < 10 ÷ 15 mm, Δy ≈ xj, |Δy|max < 3 ÷ 5 mm. The error magnitude increases with the distance from the guide axis. The bushing skews contribute mostly into the total error Δy which leads to the greater ruler skew than described above: using direct measurements made by use of a micrometric indicator it is shown that the errors reach 25 microns and more in the zone of high X values. The positioning errors (the carriage does not come out into a specified position) are related to the slackness of the power gear cog-wheels, the eccentricity of the tension pulley and the stretching of the cogged rubber belt (Figs. 3 – 5). Errors of this kind reach maximum values ranged as – 20 < |Δy|max < 30 microns in the high Y-value zone, the situation getting worse as far as the mechanical part elements aging. The values given for the astronomical plate position measurements with a scanner, more correctly speaking, for the measurements of images distorted during the scanning, are in accordance with the results obtained by other authors and by use of other techniques [7, 8]. Several attempts to use scanners for digitization of astro-plates have been undertaken since the beginning of the flatbed scanners applications. In the mid of 1990s, even a manual scanner had been used to obtain the provisory coordinates of objects on astro-plates. In spite of a fast progress of using the commercial scanning devices and the successful flatbed scanner application for solution of a narrow class of problems in the field of astrometry this technique has been considered and applied as an auxiliary equipment only because of availability of the measuring engine FANTAZIA at Pulkovo. Now we can apply e new equipment based of the progress of the development of the technologies in electronics and micro-mechanics, the of comparatively low cost positioning mini-sensors. The problem is only with the price of the sensors which exceeding the cost of a scanner itself by one order of magnitude. The cheap solution proposed is based upon application of a special bench-marked certified glass scales converted scanner into a highly precise instrument according our preliminary investigations and allow to avoid using the expensive precision sensors. For contact: E. Poliakow: THE BENCH-MARKING OF THE SCALES The problems to the use of flatbed scanners related with the astro-metric tasks and methods of their solution are known. These are, first, the method of measurements by use of the one-coordinate devices which are already known consisting the twofold scanning procedure of an astro-plate with the 90-degree turn. The results of the measurement along the precise axis being combined in the consequent processing, and, secondly, the method dealing with mixing-up the standard markers with precisely known positions into a starting image. The results of the measurements of the material under investigation have being reduced to the processing of the system of markers. Both methods have accurately been investigated in the papers [7,8]. We propose the third method based on applying two standard scales (Fig. 9). The scales are located along the edges of the image under consideration and scanned together with it. Strictly speaking, this method is an advanced version of the second approach but appears to be more efficient and to have no drawbacks of its forerunner, i.e. it would make it feasible to directly measure errors of every individual scan with no reductions of measurements using the limited number of standard markers; it will make possible to compensate in full detail for an image distortions by direct computation of the position corrections for every image element. It is assumed that the CCD-ruler is solid and non-flexible, and the position of its pixels is invariable with respect to its basis, that the distortions due to the light-detector optical system are constant, these assumptions are valid on the micron accuracy level. Hindrance sources during the measurements and reflecting on the quality of the images digitized: black plastic coming from the spring-loaded adapter supports and those of the white plastic arising from the axles of the tension pulleys (Figs. 6,7), even a real bug discovered inside the scanner (Fig.8.) Fig. 4: The electric step-motor and the power gear cog-wheels. Fig. 3: A power gear, a cogged belt drive and a guide. Fig. 5: The tightening pulley, the cogged rubber belt and the mount of the guide.