Method for adjusting an equatorial or altazimuth mount

Abstract

The present invention relates to a method for readjusting a parallactic or azimuthal mounting, comprising a device which is intended for positioning and moving a telescope with a camera and can be aligned and readjusted by means of at least one image sensor and an electromotorized controller, characterized in that the image sensor acts as a main recording sensor of the camera and at the same time as an alignment sensor and readjustment control sensor, wherein before and after a main image is taken at least one control image is taken with a shorter exposure time and these control images are compared with one another, or at least a main image itself acts as a control image and is compared with at least one previous main image, or a short-exposed control image is compared with the main image itself and the correction values for the readjustment of the mounting are determined by the image offset and the time difference of the images taken. The method is the prerequisite for easy, error-free operation of an astronomical mounting for the purpose of long-exposure astronomical photography.

Claims

1.-5. (canceled)

6. A method for adjusting an equatorial or altazimuth mount with electric motors based on data received from an image sensor, the mount carrying a telescope with a camera, wherein the image sensor serves a main recording sensor of the camera and also as an alignment sensor and adjustment control sensor, and wherein before and after a main image is taken two control image with a shorter exposure time than the main image are taken and these control images are compared to one another, or at least one main image serves as a control image and is compared to at least one previous main image, or a short exposure control image is compared to a main image, and wherein correction values for the adjustment of the mount are determined by an image offset and a time difference dt of the images taken, and wherein a correction value for a movement of axes of the mount is determined and an error in the adjustment is measured regularly.

7. The method as in claim 6, comprising:: a) taking a control image K(n); b) taking a main image H(n); c) taking a control image K(n+1); d) determining the time difference dt between the control images K(n) and K(n+1) and a subpixel accurate image offset BVx and BVy in image coordinates between the control images K(n) and K(n+1); e) determining the offset in celestial coordinates HVRa and HVDek based on BVx and BVy and the position of the image in the celestial coordinates HRa and HDek; f) calculating a new function Ra(t) and Dek(t) for the angular position of the axis as a function of time, based on the time difference dt and the offset HVRa and HVDek; and g) increasing n to n+1 and repeating steps a) to g), wherein the axes are moved according to the calculated new function RA(t) and Dek(t).

8. The method as in claim 6, wherein the image offset is calculated on the basis of an updated measurement and past data.

9. The method as in claim 6, wherein functions for the movement of axes of the mount are calculated as follows: I. movement of the Ra axis with sidereal angular velocity Vsidereal
Ra(t)=Vsidereal*t,
Dek(t)=0; II. determination of a new angular velocity on the basis of the time difference and the offset
VRaKorr=HVRa/dt+Vsidereal,
VDekKorr=HVDek/dt; III. movement of the axes with the determined functions
Ra(t)=VRaKorr*t,
Dek(t)=VDekKorr*t; IV. determination of a new angular velocity on the basis of the time difference and the offset
VRaKorr=HVRa/dt+VRaKorr,
VDekKorr=HVDek/dt+VdekKorr; and V. repetition of steps III to V.

10. The method as in claim 6, wherein functions for the movement of axes of the mount are calculated under consideration of refraction as follows: I. movement of the Ra axis with sidereal angular velocity Vsidereal
Ra(t)=Vsidereal*t,
Dek(t)=0; II. determination of a new angular velocity based on the time difference and the offset taking into account the differential refraction
VRaKorr=(HVRadeltaRaRefrakt)/dt+Vsidereal,
VDekKorr=(HVDekdeltaDecRefrakt)/dt; III. movement of the axes with the determined functions
Ra(t)=VRaKorr*t,
Dek(t)=VDekKorr*t; IV. determination of the new angular velocity based on the time difference and the offset taking into account the differential refraction
VRaKorr=(HVRadeltaRaRefrakt)/dt+VRaKorr,
VDekKorr=(HVDekdeltaDecRefrakt)/dt+VdekKorr; and V. repetition of steps III to V.

Description

DETAILED DESCRIPTION OF AN EXEMPLARY EMBODIMENT

[0030] The implementation of the method according to the invention is explained stepwise in the following by way of an exemplary embodiment of an existing system. The system consists of an equatorial mount with high-resolution rotary encoders and step motors on both axes. The motor control of the system enables the evaluation of the rotary encoders, control of the motors and precise regulation of the axes.

[0031] In this case, a digital single-lens reflex camera having a live view mode is used as an image sensor. The evaluation is performed on an additional computer (laptop) with the aid of software. The software includes integrated functions such as, for example, complete control of the digital single-lens reflex, two-star method for calculating the placement error based on the position of two stars, calculation and transfer of the correction values to the controller and image processing algorithms for the offset determination.

[0032] The mount is leveled with the aid of a bubble level and is aligned parallel to the axis of the earth by means of a compass and a setting circle. In addition, the geographic coordinates of the site are entered into the software for the calculation. Also conceivable is the simplified design of the alignment and operation by the use of inclination (leveling) and direction sensors (elevation) and a GPS sensor for the automatic determination of the location.

[0033] Alternatively, the leveling can be dispensed with by determining the difference angle of the azimuthal coordinate systems of the mount and the earth and taking this into consideration in the correction.

[0034] With the integrated GoTo function, the optics is driven to a bright star which is known from the database. With the live view of the camera and the interface to the software, the user has the option of focusing the optics through the magnification of the star. The orientation of the camera on the optics is determined with the function of the automatic camera calibration, in that the controller first pivots a monitored star with the right ascension axis to one image edge, stores the image coordinates, and then moves to the other image edge in order to calculate the orientation angle of the camera.

[0035] After the camera calibration, the controller executes the two-star method to minimize the placement error. The known star is automatically moved into the center of the image, the axis positions are calibrated by the rotary encoders and pivoted to a further known star to compare the deviation of the axis position from the star position.

[0036] With the two-star equation, the software calculates the placement error of the mount and visually displays the correction values in azimuth and elevation to the user, which can be corrected in the live image on a star. After the correction has been completed, the camera alignment is automatically rechecked and adjusted if necessary. The user can then move to a desired object, set the exposure times and number of images, and start the image cycle.

[0037] The software automatically records short control images before and after each main image, which are then compared to each other. In the first embodiment of the method, in this case, only the right ascension axis is driven sidereally. The image offset between the control images is determined by an image processing algorithm (e.g., FOCAS Automatic Catalog Matching Algorithm). The exposure time and ISO value of the control images are automatically adjusted, in which the algorithm evaluates the control image and searches for sufficient details. If there are insufficient or no details to be seen, the exposure time and/or the ISO value of the control images are automatically increased.

[0038] Thus, a subpixel-accurate evaluation of the offset is achieved in both axes, the proportion of the temporal angular change between the control images in declination and right ascension is calculated by a conversion of image coordinates into celestial coordinates using the axis positions and the parameters of the optics (focal length, pixel size, . . . ) entered thereinto.

[0039] Based on these values and taking into account the differential refraction, a local error model is created at the desired location in order to obtain the correct target position as a function of the time for the declination and right ascension axis.

[0040] The effect of the refraction is well known and can therefore be corrected in the control software. The atmospheric refraction is dependent on air pressure, temperature and height of the object. However, since only the differential refraction is necessary in this method, the height of the object is primarily relevant.

[0041] The local error model calculates the placement accuracy and transforms this back to the temporally continuing image offset for the next main exposure. A polynomial of the error curve is calculated in this case and passed to the controller in order to operate the axes with a dynamic tracking speed.

[0042] This procedure is repeated for further main images, wherein the tracking speeds of the axes are recalculated and adapted each time.

[0043] By means of this method, it is thus possible to eliminate the error of the placement and the differential refraction and to changes in position or orientation between mount and the celestial object and to enable images with longer exposure times without tracking errors.

Method without Correcting the Placement

[0044] In a further exemplary embodiment without correction of the elevation and azimuth axes, the mount is first set up and the polar axis is aligned as precisely as possible. The alignment can be performed by means of a polar finder or by compass and angle scale. The more precise the placement, the smaller the image field rotation, depending on the length of the exposure, position in the sky and the size of the image field. The horizontal leveling can be dispensed with completely since this has no effect on the tracking.

[0045] After placement, the orientation of the camera is adjusted. In this case, a star is placed in the image field and the camera is oriented in such a way that the star is horizontal during a movement of the right ascension axis.

[0046] The position of the object to be exposed is now controlled, the regulation for the right ascension is activated and started with a short-exposure control image. This is followed immediately by a long main image, which is in turn concluded with a brief control image.

[0047] Based on the two control images, the image offset is now determined by a suitable image processing algorithm. The most common method is triangulation. The algorithm searches as many as possible triangles that connect the stars and thus calculates the offset and the rotation between the reference and control image.

[0048] A further possibility for offset determination is an algorithm which operates with two-dimensional cross-correlation in the image plane. The signal of the first reference image is compared with the signal of the second control image and the maximum is searched for.

[0049] Next, the calculated offset of pixels is converted to the angular dimension. The value in this case depends on the focal length of the optics and the pixel size of the camera.

[0050] Based on the offset in the angular dimension and the temporal difference of the recorded control images, a temporal angular velocity can be calculated based on the target position and actual position.

[0051] In order to achieve a tracking which is as error-free as possible, an extended error model is created, which also takes into account the proportion of the differential refraction. This results in a function of the error curve of both axes, which can be almost completely eliminated by adapting the angular velocity of the controllers.