JPH11146268A - Film scanner and method for scanning cinefilm - Google Patents

Abstract

PROBLEM TO BE SOLVED: To surely obtain the position of a perforation hole even when there are damage and stain of the perforation hole by mutually offsetting scan value and the value of a reference pattern and generating a correction signal from an absolute offset of the scan value and the reference pattern which detect optimum equality. SOLUTION: An evaluating device mutually offsets scan value and the value of a reference pattern, also calculates the equality degree of the scan value and the reference patter in each calculation step and generates a correction signal from an absolute offset of the scan value and the reference value which detect optimum equality. Here, it is advantageous to mutually and separately calculate horizontal and vertical deviations in order to calculate the equality degree of the scan value and the reference pattern. Therefore, a dimension line 40 is provided perpendicularly to edges that are checked respectively, and for instance, the scan value 42 of an upper edge 43 is calculated from a perforation hole 41.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a scanning device for optically scanning a perforated hole, and an evaluation device for comparing a scanning value of the perforated hole with a reference pattern to generate a correction signal for compensating an image position error. And a film scanner having the same. The invention also relates to a method of scanning a cine film, which optically scans the perforation holes of the film and compares it with a reference pattern to generate a correction signal for compensating for image position errors.

[0002]

2. Description of the Prior Art In order to convert cine film material into electrical signals, the film in a film scanner passes through or near a photoelectric scanning device. In this case, there has conventionally been a problem that the image position of the sequentially scanned image must be kept constant. Partially periodic image position fluctuations and partly steady image position fluctuations are referred to as image position errors, which have various causes.
One of the possible causes is a positioning error in the photographing camera and the negative / positive copier. However, on the other hand, image position errors and tracking errors of the film scanner can cause other image position errors.

[0003] Different solutions exist to reduce image position errors. For example, from DE 37 36 789 A1, it is known to scan perforation holes, one for each of these images, as reference points for each film image to be scanned. A pulse signal is generated by a scanning device, for example, a line sensor provided at a predetermined angle to the running direction of the film, and this signal is compared with a stored reference pattern. From the temporal deviation between the pulse-like signal actually scanned in each case and the stored reference pattern, the calculation circuit detects the horizontal deviation (horizontal offset) and the vertical deviation (vertical offset) of the perforated hole scanned in each case. Are obtained as a horizontal vector signal and a vertical vector signal. These vector signals are supplied to a correction circuit in which the scanned film image is shifted relative to each other in accordance with the measured horizontal and vertical displacement of the perforated holes. Using perforation holes as reference points for film images has the advantage that mechanically constrained film positioning errors can be well assured. This is because the position of the perforated hole to be scanned in each case and the position of the film image to which it corresponds is usually correlated with very small tolerances.

However, in order to determine the position of the perforation hole by the known prior art, the geometrical shape of the perforation hole to be scanned must match the reference pattern as accurately as possible. Damage or dirt on the perforation holes is only averaged, and erroneous image position correction is necessarily performed, especially when an error occurs accidentally.

[0005]

SUMMARY OF THE INVENTION An object of the present invention is to reliably determine the position of a perforation hole even if the perforation hole is damaged or dirty. It is also necessary to increase the resolution of the film image scanned at this time, so that image position errors that were hardly visible until now can be seen. Therefore, another object of the present invention is to determine the position of a perforation hole very accurately.

[0006]

An object of the present invention is to provide an evaluation device, wherein the scanning device and the value of the reference pattern are offset from each other in a plurality of calculation steps by the evaluation device, and each evaluation is performed by the evaluation device. In the step, the degree of coincidence between the scan value and the reference pattern is determined,
The problem is solved by configuring the correction signal to be generated from a scan value at which an optimum match is detected and an absolute offset of the reference pattern. The task also involves offsetting the scan value and the value of the reference pattern in a plurality of calculation steps, determining the degree of coincidence between the scan value and the reference pattern in each calculation step, and determining a scan value and a reference value at which an optimum match is detected. The problem is solved by generating a correction signal from the absolute offset of the pattern.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS For example, a calculation unit is provided as an evaluation device, in which a scanning value of a perforated hole and a reference pattern are offset stepwise in a vertical direction and a horizontal direction with respect to each other. The calculation unit forms a criterion for matching the scan value with the reference pattern at each offset stage, and then determines an offset step at which the scan value and the reference pattern optimally match. The horizontal offset and the vertical offset based on the offset are transmitted as correction signals for compensating the image deviation.

The scan values characterizing the edges of the perforation holes are thus wholly or partly related in a manner specific to the respective perforation hole, from which the exact position or the exact position deviation is determined. Desired. By stepwise offsetting the reference pattern with respect to the scan value (or the scan value with respect to the reference pattern), an optimal match between the reference pattern and the stored scan value is determined. The values and directions of the reference patterns or scan values that must be offset to achieve an optimal match directly result in horizontal and vertical displacement of the perforated holes that are scanned each time. The best match can then be determined via a weighting function, in which case the degree of the best match, ie the quality of the match, is equal to the reciprocal of the determined match error. By selecting a correspondingly appropriate weighting function, for example by adding the square of the difference between the scan value and the reference pattern, the effects of a faulty scan value can be suppressed.

[0009] Since the quality of the match is determined for a plurality of combinations of the scanned image and the reference pattern that are offset from each other, faulty scan values are easily detected and eliminated. Since the degree of coincidence is particularly determined for each offset stage, the scanability of the scan value can be checked in each offset stage independently of the other offset stages. For this purpose, a calculation unit is preferably provided, which compares the scan values of the perforated holes calculated from the scan values, the difference of the positions of the scanned perforations being greater than a predetermined threshold value. Remove from. By forming an average from spatially adjacent measurements taking into account the expected geometry of the perforated holes, the deviation between the individual measurements and the average is determined, from which the average is determined. Strongly deviating measurements are filtered out of further signal processing. By this means, it is possible to selectively eliminate errors caused by the amount of obstacles at the film edge such as uneven illumination of the object to be measured, dirt, scratches, film damage and the like. Since multiple scan values are used, it is ensured that a sufficient number of evaluable measurement results remain in any case.

For this purpose, for example, an average value for a deviation between the scan value and the reference pattern is formed from a plurality of scan values, and a scan value having a deviation from the average value exceeding a threshold is selectively removed from the average value forming section. Further, it is configured to form a new average value from the remaining scan values. This process is repeated until all remaining scan values do not exceed the predetermined threshold. By this means, scan values that are too far from or too close to the reference pattern compared to the average of the scan values and whose difference from the reference pattern does not match other scan values are completely eliminated. By monitoring such prosability, it is sufficient to add the difference between the scan value and the reference pattern as a weighting function.

[0011] Advantageously, when determining the degree of coincidence between the scan value and the reference pattern, for the scan value corresponding to the horizontal edge of the perforation hole, only the difference in the vertical direction with respect to each horizontal reference edge is included in the comparison result. For the scanning value corresponding to the vertical edge of the perforation hole, only the difference in the horizontal direction is included in the comparison result. In this case, the weighting functions are treated as individual components of a set of values. This is because the vertical and horizontal components of the correction signal are determined completely independently of each other. Even if one of the two components of the correction signal is erroneously obtained, there is no adverse effect on the calculation of the other components.

In another embodiment of the invention, a weighting function is used to determine one value each for each lateral edge and form an average value for each opposing edge. Thereby, the center of the perforation hole is actually detected. By averaging the pairs of opposing edges, the accuracy of the position detection of each perforation hole is improved.

The computing unit preferably has a straight section as a reference pattern, which is arranged at right angles to the edge of the perforation hole to be compared in each case. In a perforation hole in which the horizontal edge is formed in an arc shape, the straight portion of the reference pattern is arranged at right angles to the tangent of each arc-shaped curved portion of the horizontal edge. The advantage of using the straight part as the reference pattern is that the position of the perforation hole is compared with the geometrical shape (straight line, arc) and the comparison is made with very high accuracy. By using the straight part of the reference pattern, measurements due to dirt or damage can be particularly easily removed. This is because scan values that are not on the straight line are not calculated from the beginning. Since the scanning value on the linear portion of the reference pattern can uniquely correspond to a predetermined point on the contour of the perforation hole, the difference between each point and the contour of the perforation hole can be obtained. This makes it possible to easily identify a scan value whose difference strongly deviates from a plurality of other scan values.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to embodiments.

FIG. 1 shows a schematic view of a main part of a film scanner as an embodiment, and shows a cooperative operation of a detection circuit for detecting perforation holes and a film scanner constituted according to the present invention. In this case, the film 1 is continuously fed from the supply reel 2 to the take-up reel 5 via the capstan roller 4 in the direction of arrow 3. The supply reel 2, the capstan roller 4, and the take-up reel 5 are electric motors 6,
These electric motors are driven by a feed control device 9, that is, a travel control device. For this purpose, the film tension is measured using the first sensing lever 11 arranged near the supply reel 2 on the film traveling path. The measured film tension is supplied to a feed control device 9, which controls a supply reel motor 6 by an internal control circuit so as to ensure as constant a film running as possible. Using the second sensing lever 12 arranged near the take-up reel 5, the film tension is similarly controlled by the take-up reel motor 8 during the take-up process. A plurality of guide rollers 10 are provided on the film traveling path, and the film traveling path of the film 1 is predetermined by these guide rollers.

For optical scanning of the film 1, the film 1 is guided by pivotable guide rollers 13 so that the film partially winds around the capstan rollers. Thus, the film feed speed is determined exclusively by the capstan roller 4 or the capstan motor 7.
In the fast-forward or fast-rewind operation, the pivotable guide roller 13 is pivoted outward from the film 1 in the direction of the second arrow 14 so that the film no longer winds around the capstan roller 4. In this case, film feeding is performed exclusively by the supply reel motor 7 and the take-up reel motor 9.

A sprocket 15 fixedly connected to an octopus disk 16 is provided as a speed sensor on the film traveling path in order to control the feed speed during scanning. The tach disk 16 sends out a pulse signal (hereinafter referred to as tacho pulse) corresponding to the feed speed, and this pulse signal is supplied to the feed control device 9. A control signal for driving the capstan motor 7 is generated depending on the parameters set on the operation panel 17 and the tach pulse sent from the tach disk 16.

At the time of feeding the film, the film 1 travels through the scanning unit 18. In this scanning unit, the film 1 is illuminated by an illumination device, which is schematically shown as a lamp 19 and a condenser lens 20. Light for projecting the image content of the film 1 is transmitted to the image scanning sensors 21 and 2.
Scanned via 2, 23, 24. The first image scanning sensor 21 is used for scanning the luminance signal W, and the other image scanning sensors 22, 23, and 24 are used for scanning the color component signals RGB. The luminance signal W and the color component signals RGB are supplied to a video signal processing unit 25, and the signals received by the video signal processing unit are respectively converted into a desired video signal format, and can be used in a studio standard manner at the interface 26. .

When the film 1 is photoelectrically converted in the scanning unit 18, the film further passes through a device for detecting an image position error. In this embodiment, the image position error detecting device, that is, the centering error detecting device, includes a second lighting device 27 on one side of the film 1, a perforation hole sensor 28 for a sprocket on the other side of the film 1, and a computer calculation unit 29. It is configured. For example, a line camera that can be added as an independent product is suitable as the perforation hole sensor 28. Further, a CCD line sensor can be used as the perforation hole sensor 28. Computer computing unit 29 evaluates the output signal of perforation hole sensor 28 and generates a correction signal C for horizontal and vertical image position errors. This correction signal is supplied to the video signal processing unit 25. In a known manner, the individual video pictures are offset correspondingly in the horizontal and vertical position by the correction signal C during the signal conversion, each time correcting the picture position error. In the film scanner described in this embodiment, a so-called digital image effect device (D
VE) is integrated. However, as a technically equivalent solution, the video signal and the correction signal may be supplied to an externally connected DVE device via an external interface.

The film advance controller 9 advantageously evaluates the tach pulses of the tach disk 16 so that substantially the same vertical pixel resolution is obtained at different speeds so that the line of the perforation hole sensor 20 can be obtained. Adapt the frequency to the selected film speed.
In order to uniquely assign a predetermined perforation hole (for example, the uppermost perforation hole) uniquely to the individual film image, in this embodiment the film scanner is characterized by the start of a new film image for another purpose. For this purpose, the respectively generated sync pulse "frame film start" is used.

Since the perforation holes corresponding to one film image are usually formed in one punching process, the geometric size of the perforation holes corresponding to each individual film image is constant. Unless there is a very badly deformed or damaged perforation hole, it is possible to calculate a sufficiently accurate reference point for each image by scanning the correspondingly located perforation hole for each image. is there. Scanning and evaluating the perforation holes or all perforation holes corresponding to one film image indicates that if there is a perforation hole having significant damage, the perforation holes or all perforation holes corresponding to one film image may be detected. Improvements can be made to determine the reference point by averaging.

In this embodiment, in order to scan the perforation hole, the perforation hole is illuminated by using an infrared light source, for example, an infrared diode, and scanning is performed by an infrared-sensitive line sensor. Film 1 to be scanned
An infrared cut filter 30 is arranged between the image scanning sensors 21, 22, 23, and 24, and the cut filter has transparency in a visible light region. In this way, obstacles to the image scanning sensors 21, 22, 23, 24 due to light used to illuminate the perforation holes are avoided.

By illuminating the perforation holes with infrared light, the edges are surprisingly particularly easy to detect because the obtained scanning signal has a maximum of attenuation in the edge region of the perforation holes. The reason for the maximum attenuation may be that the punching process causes a change in the permeability of the material in the infrared transmitting region in the edge region. FIG. 2 shows a cross-sectional view of the film material 1 cut along the longitudinal axis, the film material having a normal transmissive area 31 and an area 32 of reduced transmissivity due to the punching process. And a region 33 of a perforation hole. The output voltage U _s of sprocket hole 29 on the lower side are indicated by the same position along the feeding direction of each film. A change in the course of the signal is preferably detected by a differentiating circuit. FIG. 2 shows the peaks 34 and 35 of the relevant pulse-like signal obtained by the gradient method. These signal peaks are supplied to the computer calculation unit 29 as scan values.

Advantageously, the perforation holes are arranged slightly inclined with respect to the horizontal axis of the film to be scanned, so that the scanning detection accuracy is improved. FIG.
Similarly to the scanning pattern of the perforation hole 36 shown in FIG. 7, the geometrical distortion of the detected scanning value occurs in the line scanning. However, this geometric distortion is because the time deviation is proportional to the film feed speed,
It can be eliminated by simple conversion of the scanning value. All scan values of the perforated holes are calculated by computer
At 9 (not specifically shown), it is temporarily stored in a memory. The evaluation device can randomly access the scan values temporarily stored in the memory.

Next, the steps of an appropriate method will be described with reference to the flowchart shown in FIG. 5 in order to determine the optimum degree of coincidence between the reference perforation hole and the scan value of each perforated hole scanned. Although the scan values can usually be compared and evaluated with reference points of a given reference image, it has proven advantageous to calculate the horizontal and vertical offsets separately from each other. For this purpose, dimension lines are used instead of reference perforation holes formed from reference points. The dimension lines are each provided at right angles to the edge to be examined. FIG. 4 shows a reference pattern formed from such dimension lines 40 and a contour of the perforation hole 41 based on the reference pattern. FIG. 4 also shows the scan values 42, 44, 46 and 48 of the perforated holes that have been further scanned. For simplicity, a scan value 42 corresponding to the upper edge 43 of the perforated hole to be scanned and a scan value 4 corresponding to the left edge 45 of the perforated hole to be scanned, respectively.
4. Right edge 47 of the perforated hole to be scanned
And only the scan value 48 corresponding to the lower edge 49 of the perforated hole to be scanned are shown.

For clarity, only a few dimension lines are shown. The meaningful maximum number of dimension lines provided for evaluation is determined by the cost of the scanning unit and the computer unit of the perforation holes. In this embodiment, the basic stability of about 30 pixels of the splice is obtained by means of corresponding measures, for example, the support guide of the film edge in the scanning unit. Advantageously, the scan values obtained at the corners of the perforation holes are also filtered out of the evaluation. This is because the evaluation of the curved portion of the corner requires a relatively large calculation cost, and conversely, a large number of scan values remain without the evaluation, so that no serious adverse effect on the accuracy is caused.

In order to reduce the number of times the comparison between the reference pattern and the respective scan value has to be performed, the search is advantageously refined in stages starting from a coarse search. In a first calculation step 501 for initializing the search process, a step width n for offsetting the reference pattern is first set to eight pixels. The positions x and y where the search is started are set to zero positions x = 0 and y = 0 of the reference perforation hole at the time of initialization.

In the next calculation steps 502 to 510, a matrix E [i, j] is formed.

At -4≤i≤3 and -4≤j≤3, E [i, j] = F (SrcWin (x + ni *, Y + n *)
j); RefWin (x, y)) Each possible offset position i, j in the search window
Has a comparison error E between the scan value and the reference pattern. To calculate the matrix E [], the scanning values of the perforated holes are started at the lower left corner of the search window (i = -4, j = -4) and left to right by two interlacing counting loops. Offset to In this case, the counting loop consists of the running variables i, j
Incrementation block 50 for
5, 507 and decision blocks 506, 509 for determining the end of each loop. In each step, the offset scan value SrcWin () and the reference pattern RefWin () are calculated using the weighting function F () of the calculation step 502.
Is calculated, and the matrix E is calculated where appropriate.
Is assigned to [i, j]. When the right boundary of the search window is reached (i = 3), the search window is offset by one position upward (j = i + 1) and a new search on the left side of the search window is started. The calculation of the matrix E [] is terminated when the upper right corner (i = 3, j = 3) of the search window is reached.

As the degree of coincidence between the reference pattern RefWin and the scan value SrcWin increases, the comparison error E generated during the comparison becomes smaller. Since the horizontal displacement and vertical displacement of the perforation holes in this embodiment is mutually calculated separately, matrix elements E [i, j] with respect to the value set e _x, e _y occurs. The first value e _x Comparison error for comparison of the vertical edges of the set value, that represents the comparison error in a horizontal direction, compared error second value e _y for comparing the horizontal edge, i.e. the vertical direction of the comparison error Represents For this purpose, on the one hand, for all scan values k = (1... K) of the upper and lower edges of the perforation hole above the vertical dimension line of the offset reference pattern, respectively, the distance d _y (k) (see FIG. 4) is calculated for the upper and lower edges of the offset reference pattern. Since the upper and lower edges of the reference pattern are linear, the distance d _y (k) is equal to the vertical offset between the respective scan value 42 and the upper or lower edge of the reference perforation hole 41. Similarly, on the other hand, the distance d _x (see FIG. 4) is calculated for the scan value 1∈ (1... L) corresponding to the lateral edge. This distance d
_x is the scanning point 44 located on the dimension line
Is the distance to the intersection of the dimension line and the contour of the reference pattern. The resulting sum is divided by the number K or L used in each calculation, and the comparison error averaged by this sum

[0031]

(Equation 1)

Occurs. In culling step 503 following the calculation step 502, each average value _e x, e
All scan values whose absolute difference from _y is greater than a predetermined threshold are _culled from the actual scan values. Calculating the comparison error step 502 as long as it is still determined in the subsequent decision block 504 that the scan value has been filtered out.
Is repeated. In this case, the scan values that have been rejected in the reject block 503 are no longer considered. The calculation step 502 and the culling step 503 for scan values that deviate very strongly from the average are repeated until there are no more scan values to cull. In this way, scan values resulting from dirty or damaged edges can be reliably filtered out without adversely affecting the comparison result. In this way, erroneous decisions are avoided.

As soon as the matrix E [] is formed, in the next calculation step 510 the matrix element of the set of values with the least error is searched. Due to the address of this matrix element, a new search area x, for a new calculation process with a reduced search width n = 1
y is determined. After all of the calculation steps 502 to 509 have been repeated once more, the matrix element with the smallest comparison error represents the deviation to the accuracy of one pixel of the perforation hole sensor. Since a plurality of scan values are used, the search width is again reduced to n = 1/8 and the calculation steps 502 to 509 are newly performed, so that the resolution is higher than the predetermined resolution of the perforation hole sensor. Misalignment is required with accuracy. From the remaining reference points, the corresponding geometrical shape "straight line" or "circular arc" is approximately determined, from which the sub-pixel accuracy deviation is determined.

[0034] For sprocket hole having an arcuate transverse edges, and the scanned points are arranged on the dimension line, the calculation of the lateral distance d _x between the intersection of dimension line and the reference pattern in other calculations It is relatively troublesome in comparison. This is because the trigonometric function must be calculated. However, as shown, it is possible to use dimension lines (see FIG. 6) that lie on orthogonal scanning rasters instead of dimension lines that are perpendicular to the tangent of each arc. The result can be obtained.

In another embodiment, the position of the perforation hole is detected in two stages in a perforation hole having an arcuate lateral edge (for example, known as an "N" type perforation hole). That is, in the first stage, as described above, the positions of the upper and lower edges of the perforation hole are detected separately from each other,
An axis of symmetry between the two edges is determined. In the second stage, two arc-shaped lateral edges are correlated with a complete circle centered on the axis of symmetry already determined. The center of the circle obtained at the time of correlation between the edge and the circle is the same as the center of the perforation hole to be obtained.

Further, various different applications are possible. Means may be used that does not calculate the center of the hole and uses a portion of the edge of the perforated hole directly as a reference.

[Brief description of the drawings]

FIG. 1 is a schematic diagram of a film scanner.

FIG. 2 is a diagram illustrating a waveform of a perforation hole and a signal related thereto in an infrared region.

FIG. 3 is a diagram showing scan values obtained by a perforation hole sensor tilted with respect to a horizontal axis of film feeding.

FIG. 4 is a diagram showing a reference pattern arranged at a right angle to an edge of a perforation hole in a perforation hole having an arc-shaped lateral edge.

FIG. 5 is a flowchart for explaining calculation steps performed by a calculation unit.

FIG. 6 shows another reference pattern for a perforation hole having an arcuate lateral edge.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Film 2 Supply reel 4 Capstan roller 5 Take-up reel 6, 7, 8 Electric motor 9 Feed control device 10, 13 Guide roller 11, 12 Sensing lever 15 Sprocket 16 Tach disk 17 Operation panel 18 Scanning unit 19, 20, 27 Illumination devices 21 to 24 Image scanning sensor 25 Video signal processing device 26 Interface 28 Perforation hole sensor 29 Calculation unit 30 Infrared cut filter 31 to 33 Each region of film 34, 35 Signal peak 36, 41 Perforation holes 37, 42, 44, 46 , 48 scan value 40 dimension line 43, 45, 47, 49 edge of perforation hole

 ──────────────────────────────────────────────────の Continued on the front page (71) Applicant 590000248 Groenewoodsweg 1, 5621 BA Eindhoven, The Netherlands

Claims (9)

[Claims] 1. A film scanner comprising: a scanning device that optically scans a perforation hole; and an evaluation device that generates a correction signal for compensating an image position error by comparing a scanning value of the perforation hole with a reference pattern. An evaluation device (29), by which the scanning value and the value of the reference pattern are offset from each other in a plurality of calculation steps, and by the evaluation device, the scanning value and the reference pattern in each calculation step. And a correction signal is generated from a scanning value at which an optimum match is detected and an absolute offset of the reference pattern. 2. An evaluation device (29) is provided which horizontally and vertically offsets a scanning value of a perforation hole and a reference pattern with respect to each other in a stepwise manner. A criterion for matching the values is formed, an offset step at which the scan value and the reference value optimally match is obtained, and a horizontal offset and a vertical offset based on the offset step at which the scan value and the reference value optimally match are determined. The film scanner of claim 1, wherein the offset is used as a correction signal to compensate for image position errors. 3. An evaluation device (29) is provided by which the horizontal offset and the vertical offset determined in the comparison process are performed with reduced step width as a starting point for a further comparison process. The film scanner according to claim 1. 4. An evaluator (29) is provided, the evaluator comprising, if the difference calculated from the scan value to the position of the scanned perforation hole exceeds a predetermined threshold value. 4. The film scanner according to claim 1, wherein the scanning values of the perforated holes are filtered out of the comparison process. 5. An evaluation device (29) is provided, the evaluation device forming an average value from a plurality of scan values with respect to a deviation between a scan value and a reference pattern, and a deviation from the average value exceeding a threshold value. 5. The film scanner according to claim 4, wherein the scanning values to be performed are selectively removed from the average value forming section, and a new average value is formed from the remaining scanning values. 6. The evaluation device according to claim 1, wherein the evaluation device has a plurality of dimension lines, each dimension line being approximately perpendicular to the edge of the perforation hole to be compared. Any one
The film scanner according to the item. 7. An evaluation device (29) is provided, by which the horizontal component of the correction signal is formed only from the scanning values corresponding to the lateral edges of the perforation holes, and the vertical component of the correction signal is perforated. 7. A film scanner according to claim 1, wherein the film scanner is formed only from scan values corresponding to the upper and lower edges of the hole. 8. An evaluation device (29) is provided, by which the deviation of two edges at opposing edges is calculated separately from one another and the center line of the separately calculated edges is deviated. The film scanner according to claim 7, which is used as a scanner. 9. A method for scanning a cine film, comprising: optically scanning a perforation hole of a film to compare with a reference pattern to generate a correction signal for compensating for an image position error. Offset the value and the value of the reference pattern from each other, determine the degree of coincidence between the scan value and the reference pattern in each calculation step, and calculate the correction signal from the scan value at which the best match is detected and the absolute offset of the reference pattern. Generating a cine film.