JPH0519749B2 - - Google Patents

Description

[Detailed description of the invention]

<Industrial Application Field> The present invention relates to a method of electrically creating a video signal of an image rotated by 90 degrees with respect to an original image by using a digital memory. <Prior art> Recently, it has become possible to image negative film with an imaging device, record the resulting video signal on a disk-shaped magnetic sheet as a still image, and display the image on a television receiver or monitor when necessary. A system for viewing this has been developed. In this case, the original image on the negative film is in various positions depending on the frame, such as upright, horizontal, and inverted, but it is generally necessary for the image to be displayed on a television receiver in an upright position. Conventionally, therefore, a method has been adopted in which an image rotating mirror device is used to optically rotate the original image by 90 degrees or 180 degrees, and the rotated image is captured. However, image rotation mirror devices are extremely expensive. Therefore, a method of obtaining a video signal of an image rotated by 90 degrees or 180 degrees by temporarily storing the video signal in a digital memory, converting the address and reading it out is considered. That is, as shown in FIG. 2, the memory elements are arranged in [m+1] columns×corresponding to the pixel arrangement.
Considering the digital memory 2 arranged in a matrix of [n+1] rows, first, the raster scanning video signal is A/D converted, and the pixel data PD _i0 to PD _io of the i-th horizontal scan on the screen is converted to All the pixel data PD ₀₀ -P _no for one screen are stored in the digital memory 2 by sequentially writing into the i-th storage elements M _i0 -M _io as shown in FIG. After that, if the original picture is oriented horizontally to the left as shown in Figure 3a, PD _n0 ~ PD ₀₀ → PD _n1 ~ _{PD 01}
→...→PD _no ~ PD _0o , the pixel data is read out vertically from the 0th column to the nth column from bottom to top as shown by arrow V. Further, D/A conversion is performed as necessary. As a result, a video signal of an image rotated by 90 degrees with respect to the original image can be obtained. In addition, in the case of an original image facing right, pixel data is read out from top to bottom in the vertical direction from the nth column to the 0th row, such as PD _0o to PD _no →...→PD ₀₀ to PD _n0 . If the original image is inverted, pixel data is read out horizontally from right to left from the m-th line to the 0th line, such as PD _no to PD _n0 →...→PD _0o to PD ₀₀ . If the original picture is erect,
Of course, it is sufficient to read out the data in the same order as when writing, that is, from the 0th line to the mth line in the horizontal direction from left to right. <Problems to be Solved by the Invention> However, when obtaining a video signal of an image rotated by 90 degrees with respect to the original image, generally the sampling frequency f _S of A/D conversion for recording the video signal in a digital memory is If the clock frequency _fck for reading out the video signal from the digital memory is matched, the original image 12 in Figure 3a becomes the image 13 in Figure 3b, and the horizontal and vertical dimensions of the image change. There is an inconvenience. The reason is as follows. Where, f _H : Horizontal scanning frequency (Hz) N: Number of horizontal scanning lines (lines/frame) A: Horizontal dimension of effective screen of image display device B: Vertical dimension of effective screen of image display device R _asp : Image display Aspect ratio of effective screen of device (R _asp = A/B) η _H : Effective horizontal scanning area ratio of image display device (=
A/X) N _H : Number of effective horizontal scanning lines of image display device X: Horizontal dimension of ideal screen (=A/η _H ) Y: Vertical dimension of ideal screen (=N・B/N _H ) . However, the effective screen refers to a portion of all video signals in which a luminance signal (RGB in the case of color) exists, excluding the horizontal retrace period, vertical retrace period, and so-called blanking period. Furthermore, an ideal screen is a screen in which no effective screen is missing from the original image when the original image is rotated, and the aspect ratio does not change. Now, suppose that a video signal is A/D converted at a sampling frequency _fS , written to a digital memory _at a clock frequency fS, and read out at a clock frequency _fS in the same address order as written. , the horizontal pitch P _H of the pixel is: P _H =X・_H /f _S = A・f _H /η _H・f _S (mm) ...Equation (1) The vertical pitch Pv of the pixel is Y/N=B/N _H (mm) ...Equation (2). On the other hand, if the pixels are rotated by 90 degrees and read out along the vertical direction at the same clock frequency of f _S as sampling, the pixels that were arranged vertically at B/N _H pitch will now be A・f _H /η _H・They will be lined up horizontally at f _S pitch, and the horizontal magnification M _H will be M _H = A・f _H /η _H・f _S /B/N _H = A/B・N _H・f _H /η _H・
f _S ...Equation (3) is obtained. In addition, pixels that were arranged horizontally at A・_H /η _H・_S will be arranged vertically at A/N _H pitch, so the vertical magnification M _V becomes M _V = B/N _H / A・f _H /η _H・f _S =1
/M _H =η _H・f _S /A/B・N _H・f _H ...Equation (4) is obtained. In this case, it is practically difficult for M _H =M _V =1. That is, in a normal television system,
A/B=4/3, N _H ≒442, η _H ≒0.776, _H = 15.75KHz
Therefore, in order to make M _H = M _V , it is necessary that f _S ≒12 MHz. However, with a sampling frequency of 12 MHz, it is impossible to faithfully sample a video signal having a band of 0 to 7 MHz. Therefore,
For example, if the video signal is sampled at 14MHz, and the readout clock frequency required to obtain the video signal of an image rotated by 90 degrees with respect to the original image is _fck , then B/N _H :A・f _H /η _H・f _S = A・f _H /η _H・f _ck :B/N
_H ...Equation (5) ∴f _ck = (A/B) ²・N _H ²・f _H ² /η _H ²・f _S ...Equation (6) must hold. Therefore, in general, f _ck ≒ (4/3) ²・(442) ²・(15750) ² /14×10 ⁶
×(0.776) ² ≒10.2MHz, and f _ck and f _S do not match. Thus, the mismatch between the sampling frequency f _S and the readout clock frequency f _ck makes system design extremely complicated. In view of the problems of the prior art described above, the present invention provides a method in which the sampling frequency and readout clock frequency match when creating a video signal of an image rotated by 90 degrees using a digital memory. With the goal. <Means for Solving the Problems> An image signal rotation method according to the present invention that achieves the above-mentioned object rotates a raster scanning video signal to a frequency approximately equal to √K・A/B・N _H・f _H /η _H The sampled pixel data is stored in a digital memory as pixel columns in each row in the horizontal direction for each pixel data of one horizontal scan, and the stored pixel data is stored along the pixel columns in the vertical direction at a cycle time of 1/f. [K-1] It is characterized by reading out every other column. However, f _H is the horizontal scanning frequency, N _H is the number of effective horizontal scanning lines of the image display device, A/B is the horizontal to vertical dimension ratio of the effective screen of the image display device, and η _H is the effective horizontal scanning area ratio of the image display device. = horizontal dimension of effective screen of image display device/horizontal dimension of ideal screen, K is an integer of 2 or more. Also, as mentioned above,
The effective screen refers to a portion of all video signals in which a luminance signal (RGB in the case of color) exists, excluding the horizontal retrace period, vertical retrace period, and so-called blanking period. Furthermore, an ideal screen is a screen in which no effective screen is missing from the original image and the ratio of height and width does not change when the original image is rotated. <Function> Even if the sampling frequency and readout clock frequency are equal, if pixel data is read out every [K-1] columns in the vertical direction, the vertical magnification on the screen M _V (K) is 1/K of the magnification M _V (formula (4) above) when reading out one column at a time. That is, M _V (K)=1/K・M _V =1/K・η _H・f _S /A/B・N _H・f _H
...Equation (7) So, if we find the sampling frequency f _S that satisfies M _V (K) = M _H , we get: 1/K・η _H・f _S /A/B・N _H・f _H = A /B・N _H・f _H /
From η _H・f _S , _S = √K・A/B・N _H・f _H /η _H (Hz) ...Formula (8) M _V (K)=M _H =1/√K ...Formula ( 9) becomes. Based on the above, the sampling frequency and readout clock frequency are set to √K・A/B・N _H・f _H /η _H , and the pixel data is
1] By reading every other column, it is possible to obtain a video signal of an image rotated by 90 degrees with respect to the image before rotation and reduced to 1/√K both vertically and horizontally. However, it is not necessary that the sampling frequency and the readout clock frequency exactly _{match √K.A/B.NH.f H /} η _H. <Example> An example of the present invention will be described with reference to FIGS. 1 to 4. FIG. 1 shows an example of the configuration of an apparatus for implementing the method of the present invention. In the figure, 2 is a digital memory, 3 is an address generator, 4 is a clock generator, and 5 is an A/D.
The converter 6 is a D/A converter, and the image rotation device 1 is constituted by these converters. 7 is an imaging device, 8 is a subject, 9 is an image display device, 10 is a recording device, and 11 is an operating device. In FIG _{. 1, a clock generator 4 generates a clock pulse 4a having a frequency of f=√K.A/B.NH.f H /} η _H (Hz) given by equation (8).
For example, if K=2, A/B=4/3, N _H =442, f _H =15.78 KHz, and η _H =0.776, then f≈16.9 MHz. The A/D converter 5 is controlled by the clock pulse 4a, samples the video signal 7a from the imaging device 7 at a frequency of √K・A/B・N _H・f _H /η _H , and converts the digitized sample value (pixel data) 5a to digital memory 2
give to The digital memory 2 is connected to the A/D converter 5 by the R/W command signal 11a from the operating device 11.
, or read out stored pixel data. In this case, the write address and read address of the digital memory 2 are specified by the signal 3a from the address generator 3. The address generator 3 is controlled by a clock pulse 4a and updates the address every η _H /√K·A/B· _NH ·f _H (seconds). Further, the address generator 3 receives an image rotation mode setting signal 1 from the operating device 11.
1b to output an addressing signal corresponding to one of the following four modes, for example. (1) Right 90° rotation mode for erecting a left horizontal image: Considering the digital memory 2 shown in Figure 2, when writing, the i-th horizontal scanning pixel on the screen of the raster scanning video signal is It is assumed that data PD _i0 -PD _io are sequentially written into the i-th storage elements M _i0 -M _io from the beginning as indicated by arrow H, and all pixel data PD ₀₀ -PD _no for one screen are written. Then, in this 90° clockwise rotation mode, ``If we first read out the pixel data in the 0th column in the direction of arrow V, as in PD <sub>n0</sub> → PD <sub>00</sub> , then the pixel data in the Kth column will be read from bottom to top → 2K column pixel data from bottom to top→…→qK column pixel data from bottom to top'', the vertical pixel data rows are sequentially read from the left every [K-1] columns from bottom to top. The address will be updated as follows. (2) 90° left rotation mode for erecting a right landscape image: In this 90° left rotation mode, ``First, read out the pixel data in the nth column from top to bottom as PD _0o → PD _no . Then, from then on, the pixel data of the n-K column goes from top to bottom → the pixel data of the n-2K column goes from top to bottom →... → the pixel data of the n-qK column goes from top to bottom.'' The addresses are updated so that the vertical pixel data rows are sequentially read from top to bottom every [K-1] columns from the right side. (3) 180° rotation mode to make an inverted image upright: In this 180° rotation mode, addresses are updated so that the horizontal pixel data rows are read out sequentially from the bottom and from right to left, contrary to the writing process. do. (4) Non-rotation mode when there is no need to rotate the image: In this non-rotation mode, addresses are updated in the same order as they were written. In addition, in both left and right 90° rotation modes,
The interval between reading out a certain vertical pixel data string and reading out the next vertical pixel data string is adjusted to the horizontal scanning period (1/f _H ). In this case, if a left and right margin in FIG. 3B or a margin 15 as shown in FIG. 3C occurs during rotation, address updating is stopped for the time corresponding to the margin. Further, the digital memory requires only the minimum capacity necessary for the effective screen, and therefore, address updating and read/write operations are stopped during the horizontal retrace period, the vertical retrace period, and the so-called blanking period. However, even if address updates are not stopped, writing and reading of digital memory is performed in a completely one-to-one correspondence for one frame of images at the pixel level, and different columns are in the same horizontal column. It is never output to . If the video signal is of the interlaced scanning type, the pixel data is divided into fields in the vertical direction so that when rotated by 90 degrees, vertical pixel data rows are read out every [K-1] rows. This is considered to be an update of the read address. Also, when storing one field's worth of video signals in digital memory,
The number of effective horizontal scanning lines in one field is used as N _H. Further, when rotated by 90 degrees, the image 12 in FIG. 3a is reduced to 1/√K as the image 14 shown in FIG. 3c, and a blank area 15 is created on the screen. Therefore, so that this blank area 15 has an appropriate background color and background pattern,
It is preferable to add an appropriate signal to the video signal read from the digital memory 2 and created. Furthermore, the value of K will be described. Considering the case of displaying a negative film image on an image display device,
While the aspect ratio of negative film is generally 5:3, the aspect ratio (A/B) of the screen of a general television receiver or monitor is 4:3. Therefore, the negative film image 17 is often displayed horizontally on the television screen 16 as shown in FIG. 4a. In such a case, if the image is rotated by 90 degrees, a vertically elongated negative film image 18 will be displayed, reduced to 1/√K, as shown in FIG. At this time, if we set K=2,
The negative film image 18 is not too small and does not protrude from the screen 16, and it takes almost no effort to adjust the magnification of the imaging device. Next, a description will be given of the operation when still recording the subject 8 with the recording device 10 using the image rotation device 1 shown in FIG. 1. First, the video signal 7 from the imaging device 7
a to the image display device 9, and performs necessary processing such as focus adjustment, magnification adjustment, and color correction while viewing the displayed image. At this time, if the subject 8 is a negative film, it is preferable that the video signal 7a represents a positive image. When it is necessary to rotate the displayed image, a write command signal is given to the digital memory 2 and the address generator 3, and the A/D converter 5 rotates the image as follows: =√K・A/B・N _H・f _H /η Pixel data 5 obtained by sampling at a frequency of _H
A is written into the digital memory 2 for one screen. After that, set the image rotation mode to the desired one,
Further, a read command signal is given to the digital memory 2 and the address generator 3, and pixel data is read out from the digital memory 2 at a clock frequency of f=√K·A/B·N _H · _H /η _H. Pixel data 2a from the digital memory 2 is normally provided to the image display device 9 and the recording device 10 after being subjected to D/A conversion. The video signal 6a for one screen is repeatedly applied to the image display device 9, but the video signal 6a for one screen is repeatedly applied to the image display device 9.
It is sufficient to supply the video signal 6a for one screen. The video signal 6a given to the recording device 10 is transferred to an appropriate recording medium 10a such as a disk-shaped magnetic sheet.
recorded in Note that 11c is a recording command signal;
1d is a reproduction command signal. The video signal recorded on the recording medium 10a is reproduced as necessary and provided to the image display device 9 to monitor the image quality and the like. On the other hand, if image rotation is not required, the video signal 7a from the imaging device 7 may be directly supplied to the recording device 10, or
Alternatively, the video signal 6a obtained by reading out the digital memory 2 in non-rotation mode is provided to the recording device 10. By performing image rotation processing as described above, a large number of photographic images in a roll of negative film or a printed photo album or an image of an arbitrary subject can be
Regardless of the original state, it can be electrically restored to an erect image state and recorded on a magnetic medium or the like.
By applying a magnetic medium or the like to the reproducing device at any time, an erect image is displayed on the image display device, so viewing the image is comfortable. <About digital memory> By the way, considering the case where K = 2, A/B = 4/3, N _H ≒ 442, η _H ≒ 0.776, f _H = 15.75 (KHz), ≒
16.9 (MHz), so the digital memory 2 needs to be fairly high-speed with an access time of 60 ns or less. Therefore, for example, access time is slow.
A method of realizing functions similar to high-speed digital memories by using a large number of digital memories with a speed of 200 to 300 ns will be explained with reference to FIGS. 5 and 6. Figure 5 shows four digital memories 20A, 20
It is a block diagram of an example using B, 20C, and 20D. In the figure, 21 is a 4-stage shift register;
22 is a latch circuit, 23 is a distribution switch, 24 is an address generator, 25 is an address converter, 2
6 is a 4-stage shift register, 27 is a 1/4 frequency divider,
4 is a clock generator of √K・A/B・N _H・f _H /η _H (Hz). In FIG. 5, the video signal 7a is sampled by the A/D converter 5 at a frequency of √K・A/B・N _H・f _H /η _H , and the pixel data 5a
are sequentially input to the shift register 21. Similarly, a clock pulse having a frequency of √K·A/B·N _H ·f _H /η _H is input to the shift register 21. The data in each stage of the shift register 21 is transferred to a latch circuit 22.
is input. Since the shift register 21 has four stages, the frequency of the clock pulse of the latch circuit 22 is set to √K·A/B·N _H ·f _H /4η _H. As a result, four pieces of pixel data 5a from the A/D converter 5 are fetched in parallel.
The acquisition period is four times the sampling period, which is slower. Four outputs 22 of latch circuit 22
A, 22B, 22C, 22D are connected to four digital memories 20A, 20B, 20 through a switch 23.
C, 20D are given separately. Therefore, it is sufficient for each digital memory to have an access time of 4η _H /√K·A/B·N _H ·f _H or less, that is, four times the sampling period or less. On the other hand, the switch 23 connects the four outputs 22A, 22B, 22 of the latch circuit 22.
C, 22D and four digital memories 20A, 20
The connection relationship with B, 20C, and 20D is controlled by the control signal 28.
It operates to sequentially switch at the horizontal scanning period. As a result, P ₀₀ ~
Pixel data for one screen of P _no is distributed and written to four digital memories A, B, C, and D in both the horizontal and vertical directions. In FIG. 6, A indicates that pixel data has been written to the digital memory 20A, and similarly, B, C, and D indicate that pixel data has been written to the digital memories 20B, 20C, and 20D. As shown in Figure 6, if four adjacent pixel data are written separately in the four digital memories in both the horizontal and vertical directions, there will be no error when the data is rotated by 90 degrees to the left or right, or if it is rotated by 180 degrees. In any case of rotation, four digital memories 20A, 2
Pixel data of 0B, 20C, and 20D can be read out simultaneously at a clock frequency of √K·A/B·N _H ·f _H /4η _H. The four read pixel data are input in parallel to each stage of the shift register 26, and the shift register 2
By shifting the pixel data of 6 with a clock pulse of √K·A/B·N _H ·f _H /η _H , a desired video signal can be obtained. In addition, shift register 2
6, in the case of right 90 degrees rotation mode and 180 degrees rotation mode, it is shifted to the left, and in the case of left 90 degrees rotation mode and non-rotation mode, it is shifted to the right. Here, addresses will be explained. Now, the 6th
When writing pixel data PD ₀₀ to PD _no separately into four digital memories A, B, C, and D as shown in the figure, the four pixel data that the latch circuit 22 outputs at once are combined into one and the address is specified. Think about what to do. For example, as shown in Figure 6, 0
0-0, 0-1, 0- for the th horizontal scan
Generally, the latch circuit 22 is connected to the latch circuit 22 at the j-th position in the i-th horizontal scan, such as 2...0-j...,0-[n/4].
The address of the pixel data group output from i-
Suppose that it is determined as j. The addresses for writing are as follows for each digital memory: 0-0 0-1 0-2 0-3 ... 0-[n/4] 1-0 1-1 1-2 1-3 ... 1-[n/ 4] 2-0...m-[n/4] The address can be updated as follows. Therefore, the address translation device 25 outputs the address designation signal 24A from the address generation device 24 as it is. Reading in non-rotation mode is the same as updating the address during writing. When reading in 180° rotation mode, each digital memory has the following values: m-[n/4] m-([n/4]-1) m-([n/4]-2) m-([n /4]-3) (m-1)-[n/4] (m-1)-([n/4]-1) (m-1)-([n/4]-2) (m- 1)-([n/4]-3) (m-2)-[n/4]...0-0, the address may be updated in the opposite manner to the writing. In this case as well, the address translation device 25 may output the address designation signal from the address generation device 24 as is. When reading in the right 90° rotation mode, each digital memory updates a different address as shown in Table 1 below. However, consider the case where K=2. As can be seen from Table 1, if the address generator 24 generates the addressing signal 24A for the digital memory 20A, the address converter 25 changes only a small part of the addressing signal 24A to the other digital memory 20A.
Addressing signal 2 for B, 20C, 20D
It is sufficient to output 5B, 25C, and 25D. Therefore, the address translation device 25 can be easily realized using a small-capacity ROM.

[Table] On the other hand, when reading in 90° left rotation mode,
Update the addresses of each digital memory as shown in Table 2 below. However, K=2. As can be seen from Table 2, in this case as well, the address generator 24 generates an address designation signal 24A for the digital memory 20A.
If the address conversion device 25 generates a
Addressing signal 2 for B, 20C, 20D
It is sufficient to output 5B, 25C, and 25D. Therefore, an address translation device can be easily realized by using a small-capacity ROM. Note that when Table 2 is compared with Table 1, the updating of the address of the digital memory 20A in the 90° left rotation mode is exactly the opposite of the address updating of the digital memory 20D in the 90° right rotation mode. Similarly, updating the addresses of the digital memories 20B, 20C, and 20D in the 90° left rotation mode is as follows:
Digital memory 20A, 2 in right 90° rotation mode
This is exactly the opposite of the address updates for 0B and 20C. Therefore, it is not necessary to separately prepare the address conversion device 25 for the 90° left rotation mode and the 90° right rotation mode, and it is sufficient to switch and use three address conversion ROMs.

[Table] As explained with reference to Figures 5 and 6, even if the digital memories are slow, S digital memories that are independent of each other are used, and f=√K・A/B・Pixel data sampled at a frequency of N _H・f _H /η _H is serially input to an S-stage shift register and shifted at a cycle time of 1/f, and the pixel data of each stage of this shift register is input to the S/f shift register. During writing and reading, the pixel data of the S digital memories are written so that the S pixel data adjacent to each other in the horizontal and vertical directions are written to separate digital memories in the S digital memories in a cycle time of The S pixel data is read out with a cycle time of S/, inputted in parallel to an S stage (another) shift register, and this shift register is shifted with a cycle time of 1/f to create a pixel data string. As a result, it can operate as a digital memory that is substantially S times faster. Incidentally, when a large number of digital memories are used as shown in FIG. 5, it is not only possible to use low-speed memories, but also the following effects are produced. For example, in the case of f=√K·A/B·N _H · _H /η _H ≈16.9 (MHz), suppose that instead of using high-speed memories of 60 ns or less, 8 or 16 memories of 240 ns are used. Then each memory has 480ns or 960ns
All you have to do is access it every time. Therefore, by using the remaining time, while writing the video signal, it is possible to read out the video signal in any arbitrary mode during the writing, and monitor the image in almost real time. Note that there is almost no possibility that the write address and the read address will match. <Application Example> A system for still recording image information of each frame of a film on each track of a magnetic disk or the like and using the magnetic medium as an album will be described with reference to FIGS. 7 to 9. FIG. 7 is a block diagram of the system, and FIG. 8 is an external view of the entire system. In these figures, 100 is an optical system, 200
is a recording system, and 300 and 400 are casings. The optical system 100 is a system that extends from a light source 102 driven by a power source 101 to a color video camera deck 103, and has a wedge-shaped ND filter 104, an optical integrator 105, and a color correction filter 10 on its optical axis 110.
6, a lens deck 107, an optical shutter 108, etc. are arranged. 109 is ND filter 10
A motor 111 is used to insert an appropriate amount of film 112 into the filter 106 and the lens deck 10.
113 is a motor for moving the color correction filter 106 in and out; 114 is a crimp member for fixing the frame of the film 112 at a predetermined photographing position; and 115 is a crimp member 114 drive. 116 is a detection element for measuring the illuminance of the film frame, 117 is a solenoid for the film 112.
This is an orientation detection element for identifying the type and size of the piece, as well as the orientation of the image taken on the piece (erect, left sideways, right sideways, inverted). As shown in FIG. 9, the connection tape 112a of the film 112 and the edge of each frame 112b are marked with information such as type, size, and frame orientation so that the type, size, and frame orientation of the film can be identified.
Marks such as perforations 112c and notches 112d are formed at specific positions, shapes, or numbers corresponding to the orientation. Two lens systems 119 and 120 are arranged on the lens deck 107, and one of the lens systems is inserted onto the optical axis 110 by a motor 125 depending on the size of the film. The entire lens deck 107 is supported by a housing 300 so as to be movable in a direction parallel to the optical axis 110, and is driven by a motor 126. The camera head 103 is a color video camera, such as a three-tube type or a single-tube type with rotating R, G, and B filters, and outputs R, G, and B decomposed video signals 127 here. This camera head 103 is also supported by the housing 300 so as to be movable in a direction parallel to the optical axis 110.
driven by. Lens deck motor 12
6 and camera head motor 128 are driven by a motor drive circuit 129 to move the lens deck 107 and camera head 103 up and down to desired positions in response to commands from a central processing unit (CPU) 131. Shutter 108 and other motors 1 mentioned above
09, 113, 115, 125 and the film feed motor 139 are connected to amplifiers 132 to 13, respectively.
5, system bus 130 via 137-138
and is controlled by the CPU 131.
On the other hand, the output of the illuminance detection element 116 and the output of the direction etc. detection element 117 are supplied to amplifiers 140 and 14, respectively.
1 to bus 130. CPU13
The control system 142 including the control system 142 includes a memory 143 storing various data and programs, keyboards 144 and 145, and a paper tape reader 146 for inputting various data such as operator instructions and color correction. One keyboard 144 is used to input color correction instructions, negative/positive selection, density instructions, and the like. The other keyboard 145 is used to input film type, size, frame orientation, recording instructions, etc. Video signal 1 output from camera head 103
27 is the negative/positive switch 2 of the recording system 200
01 and is input to the image rotation circuit 1 either directly or via a negative/positive (NP) inversion circuit 202. The changeover switch 201 is the drive amplifier 147
It is connected to the bus 130 through the CPU 131 and controlled by the CPU 131. The image rotation circuit 1 here has R, G,
An A/D converter, digital memory, and D/A converter are used to perform image rotation processing for each B-resolved video signal.
Each has three converters. In addition, each digital memory has 16
It is configured using several low-speed memories. The image rotation device 1 is connected to the bus 130 and controls the rotation mode, etc.
Controlled by CPU 131. Image rotation device 1
The video signal from is input to the color correction circuit 203. The color-corrected video signal is input to a recording circuit 204, modulated by a predetermined recording method such as FM modulation, and supplied to a magnetic head 205. 206
is a magnetic disk, 207 is a recording track, 208
is the disk drive motor, 209 is the spindle,
210 is a motor power supply. magnetic head 205
is supported by a head feeding mechanism (not shown) driven by a drive amplifier 148, and the CPU 1
31, the magnetic disk 206 can be moved in the radial direction. The magnetic disk 206 rotates at a constant speed. As a result, the video signal supplied to the magnetic head 205 is selectively recorded on one of the plurality of recording tracks 207 on the magnetic disk 206. 211 is a video signal waveform monitor, 21
2 is a monitor for images to be recorded, 213 is a monitor for reproduced images, and 214 is a reproduction circuit.
The color correction circuit 203, recording circuit 204, and reproduction circuit 214 described above are connected to the bus 130, and are each controlled by the CPU 131 in terms of the degree of color correction, recording timing, reproduction timing, etc.
Note that, as shown in FIG. 8, the functions of the CPU 131 are distributed into two CPUs 131a and 131b. The operation of the system described above will be briefly explained. When the film 112 is placed on the film carrier 111, the CPU 131 reads the perforations 112c formed in the connecting tape 112a through the detection element 117 to identify the type and size of the film 112. Further, the positions of frames to be recorded on the magnetic disk 206 and selection data of the recording track 207 are input to the CPU 131 via the keyboard 145 as necessary. Next, the operator operates the keyboard 145 to release the shutter 108.
When opened, images of the frames of the film 112 are displayed on the recorded video monitor 212. In this case, the image rotation device 1 may update and write the video signal 127 not only for one screen but one after another, and at the same time read it repeatedly in a non-rotation mode.
Next, while viewing the monitor image, the operator operates the keyboard 145 to move the camera head 103 and lens deck 107 to achieve the correct shooting condition.
After that, move film 1 so that the desired frame is photographed.
12 is transferred. At this time, the CPU 131 identifies the direction of the piece through the detection element 117 or receives the direction data of the piece from the keyboard 145. Further, the CPU 131 positions the magnetic head 205 on a desired vacant track 207. The memory 143 stores lens systems 119 and 120 depending on the type and size of the film 112 and the orientation of the frame.
selection, camera head 103 and lens deck 10
7 on the optical axis 110, the relationship between the imaging magnification and the standard illuminance, the operation mode of the image rotation device 1, and the filter 1
Various data such as the necessity of 06 and the necessity of NP inversion are stored as a table. So the CPU
131 drives the lens motor 125 and the camera head motor 1 according to the output of the detection element 117.
28, driving the lens deck motor 126,
Driving the filter motor 113 with reference to the output of the photometric element 116, controlling the NP changeover switch 147,
Also, set the image rotation mode. Further, the operator performs the above various controls by operating the keyboard as necessary. When the frame image of the film 112 is thus correctly captured by the camera head 103, the operator operates the keyboard 144 while viewing the monitor image to perform color correction. Next, the operator operates the keyboard 145 to issue a recording instruction. As a result, the video signal for one screen (one field or one frame) is transferred to the magnetic disk 206.
It is recorded on the track 207 of. In the same way, the recording of the next frame is started. In an application example using the method of the present invention, image rotation of a photographic image can be performed electrically, and the apparatus can be miniaturized. Also, without using high-speed memory,
Since image rotation is possible using a dynamic memory with normal readout speed, the device can be manufactured at low cost. Furthermore, since image rotation is performed in real time, work such as color correction can be performed on the monitor. <Effects of the Invention> According to the present invention, when creating a video signal of an image rotated by 90 degrees using a digital memory, it is possible to match the sampling frequency and the readout frequency.

[Brief explanation of the drawing]

Fig. 1 is a block configuration diagram showing an example of an apparatus for realizing the method of the present invention, Fig. 2 is a diagram explaining the principle, Fig. 3 a, b, c is a diagram explaining the principle, and Fig. 4 a, b shows the degree of reduction. 5 is a block configuration diagram of an embodiment in which a low-speed memory is used, FIG. 6 is an explanatory diagram of its principle, FIG. 7 is a block configuration diagram of an example of a system to which the method of the present invention is applied, and FIG. 8 9 is an external view of the film, and FIG. 9 is a plan view of the film with identification marks. In the drawing, 1 is an image rotation device, 2 is a digital memory, 3 is an address generator, 4 is a clock generator, 5 is an A/D converter, 6 is a D/A converter, 7 is an imaging device, and 8 is a subject. , 9 is an image display device, 10
is a recording device.

Claims (1)

[Claims] 1. A raster scanning video signal is sampled at a frequency approximately equal to √K・A/B・N _H・f _H /η _H , and the sampled pixel data is horizontally divided for each pixel data of one horizontal scan. An image signal rotation method characterized by storing pixel data in a digital memory as pixel columns in each row in the direction, and reading out the stored pixel data every [K-1] columns along the vertical pixel columns at a cycle time of 1/f. . However, f _H is the horizontal scanning frequency, N _H is the number of effective horizontal scanning lines of the image display device, A/B is the horizontal to vertical dimension ratio of the effective screen of the image display device, and η _H is the effective horizontal scanning area ratio of the image display device. = horizontal dimension of effective screen of image display device/horizontal dimension of ideal screen, K is an integer of 2 or more. 2 In claim 1, K=2,
An image signal rotation method characterized in that A/B=4/3.