JPS6336596B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a high-resolution imaging device in which a high-resolution one-dimensional image sensor is constructed using a two-dimensionally arranged multi-element image sensor.

A one-dimensional image sensor with about 2000 elements has already been manufactured for use in high-speed facsimile, etc. For example, the μPD792D (manufactured by NEC Corporation) using a CCD accommodates 2048 elements in a width of 30 mm. However, in the case of large-screen facsimile or earth observation from an artificial satellite, it is desired to further increase the number of elements and achieve higher resolution. Increasing the number of elements in this one-dimensional image sensor is a very difficult problem even with current advanced integrated circuit technology.

In addition, as a method to obtain high resolution, a multiple array method is being considered in which sensors are arranged in two rows by shifting the arrangement of each element of a one-dimensional image sensor by 1/2 bit, and sampling is performed alternately. (See patent application No. 53-37494).

In this method, as shown in Figure 1, a pair of image sensors are fabricated on a silicon wafer of an image sensor 1, in which light receiving element parts 2 and 3 are arranged with a 1/2 bit shift from each other, and sampling is performed alternately. By doing so, it is possible to improve the spatial resolution characteristics, which had been limited by the sampling period. In this configuration, the problem of the target imaged location being different because the elements in the horizontal scanning direction are not lined up on the same line can be resolved by delaying one output data or inserting a memory circuit to align the time and then synthesize the data. This is solved by However, this method requires the manufacture of a mask pattern with a special manufacturing process to manufacture an image sensor with such a modified arrangement, and in order to further improve the spatial frequency characteristics, the number of sampling times must be increased to 3.
In order to double the size, it is necessary to newly create a set of three image sensors, each arranged with a 1/3 element spacing shifted from the other, from the mask pattern stage, which results in a lack of flexibility.

Another conventional method is to arrange n independent one-dimensional image sensors such that they are shifted by 1/n bits from each other. FIG. 2 is a layout diagram of image sensors when n=3. As shown in the figure, the spatial resolution can be improved by arranging the one-dimensional image sensors 4, 5, and 6, each shifted by 1/3 element in the horizontal scanning direction, and increasing the sampling frequency.

Figures 3a and 3b are system diagrams of the transmitter and receiver when one-dimensional image sensors are arranged with 1/3 element spacing shifted as shown in Figure 2. This example shows an example in which the receiver and receiver are installed separately. This transmitting section applies the outputs of the one-dimensional image sensors 4, 5, and 6 to an amplifier 7, a sample-and-hold circuit 8, and an A/D converter 9, multiplexes them using a multiplexer 10, and outputs them from a modulation transmitting section 11. Further, the receiving section passes through a receiving demodulating section 12 and is then separated by a demultiplexer 13 into data for each of the three image sensors. As shown in FIG. 2, this data captures images of parts that are widely separated from each other in the vertical scanning direction, so in order to treat them as data at the same position, delay circuits 14 and 15 are used to detect pixels that are shifted in the vertical scanning direction. It is necessary to synthesize the signals after delaying them by several times. Furthermore, as shown in FIG. 2, when several image sensors are independent, it is necessary to correct optical distortion and parallelism between the sensors because the imaging area becomes large. This correction is
Delay time correction circuits 16, 17 according to the distortion,
18 and then synthesized by a multiplexer 19. This output is applied to the image signal processing section 20, where D/A conversion image recording and other image processing are performed. This method has two problems: it is difficult to arrange and mount n independent image sensors with an accurate shift of 1/n elements, and as n increases, a larger imaging area is required, resulting in a decrease in resolution. There are drawbacks such as increased optical system distortion and the need for a delay circuit with a large amount of delay since the distance in the vertical scanning direction between the image sensors reaches 1000 elements or more.

An object of the present invention is to provide a high-resolution imaging device that improves the resolution of a one-dimensional image sensor by tilting the array configuration of two-dimensional image sensors in order to solve the problems of conventional methods. It is in.

The present invention will be explained in detail below with reference to the drawings.

FIG. 4 is a principle explanatory diagram showing the operating principle of the present invention. In the figure, 21 is a two-dimensional image sensor 22-1, 22-2,...23-1, 23-
2, . . . and 24-1, 24-2, . . are light receiving elements thereof, which are arranged in a grid pattern. This two-dimensional image sensor is used as a high-resolution one-dimensional image sensor with respect to the x-axis, and is inclined at an angle θ with respect to the x-axis, which is the scanning axis. Further, it is assumed that the vertical axis of this image sensor has 2 to 4 light receiving elements arranged at intervals _YP , and the horizontal axis thereof has, for example, 2048 light receiving elements arranged at intervals _XP . These intervals X _P , Y _P
is, for example, 14 to 15 μm. Normally, when manufacturing a one-dimensional image sensor, a row of planar image sensors is sliced one by one, so the two-dimensional image sensor used in this invention consists of two or more planar image sensors. 4
It can be easily manufactured by slicing every row (n rows).

As shown in this figure, when there is an inclination of angle θ with respect to the moving direction y-axis, τ=
After Y _P cosθ/v time, move to the position indicated by the dotted line. Therefore, the center position of the light-receiving element 23-1 passes through a point that is Y _P · sin θ away from the position that was the center of the light-receiving element 22-1 on the x-axis, and imaging can be performed at this position. . Similarly, after 2τ seconds, the position of the two-dimensional sensor moves to the position indicated by the rough dotted line in the figure, and the element 24-1 is positioned on the x-axis with 2Y _P sinθ
It is possible to image this position by passing through a point that is far away.

In other words, in order to increase the number of samplings in the horizontal direction by n times, set θ so that n・Y _P sinθ=X _P・cosθ ...(1), so that n elements are on the same line of the x axis. The sampling time may be determined so that the image is captured at . Further, the resolution in the vertical direction can be similarly increased by increasing the number of sampling times m. Furthermore, _if m is _set so that Imaging can be performed at Similarly, sampling will be performed for the elements 23-2, 24-2, . . . on the same line on the x-axis.

FIG. 5 shows the relationship between the imaging times of each element when X _P =Y _P and m=n=3 (that is, the tilt angle θ=18.3°). In the figure, "0" indicates that each element is capturing an image at a certain reference time t, and "1", "2", etc. are τ/3, 2τ/, respectively, from the reference time.
This indicates that each element is performing imaging at a time when 3, . . . have elapsed. In other words, on the x-axis there are 0,
As shown as 3, 6, -1, 2, 5,..., the reference time t, t+τ, t+2τ, t-τ/3, t+2τ/3
, t+5τ/3, . . .

The number of sampling times can be increased by three times. Therefore, by inserting 10 delay circuits with a unit of τ/3 from 0 to 10, the data captured at time 0 in the figure can be converted to 15 At the next timing τ/3 later, data in the column indicated by arrow "B" is obtained, which means that the number of sampling times can be increased three times.

FIGS. 6a and 6b are block diagrams of a transmitting section and a receiving section according to an embodiment of the present invention. Although this embodiment also shows a case in which data is sent from the transmitter of an artificial satellite equipped with an image sensor to the receiver of a ground station, the transmitter and the receiver may be integrated. First, the image sensor 21 of the transmitter
Since the system transmits the planar data imaged at a predetermined imaging time during the next imaging time according to a predetermined clock and a predetermined order, the transmitting section includes one system of amplifier 25, sample hold circuit 26, and A/D.
A converter 27 is sufficient. Therefore, Figure 3a
It has a more concise structure than that of .

On the other hand, the receiving section receives the planar data sent from the transmitting section and converts it into one-dimensional data using a delay circuit (memory). First, the receiving demodulator 12
The output is separated into three systems by the gate pulse of the demultiplexer 13, the vertical axis is corrected by the delay circuits 28 and 29 with delay amounts 2τ and τ, and then the vertical axis corrected data is synthesized by the data synthesis/separation circuit 30. Furthermore, each data block is separated into three data blocks. These three pieces of data are processed by delay circuits 31 to 34 for each delay amount τ/3, τ/3, 2τ/3, 3τ/3,
A horizontal axis (inclination) correction of 4τ/3 is performed, and these correction data are combined in a combining circuit 35, and processing such as display and recording is performed in a data processing unit 20.

The delay circuits 28 and 29 that perform this vertical axis correction are as follows:
The output of light receiving element 24-1 is 2τ, and the output of light receiving element 23-1 is
The outputs of the vertical axis elements 22 and 22 are delayed by τ, respectively.
-1, 23-1, 24-1 data are converted to data on the primary scanning axis (x axis). With this correction, the vertical axis data of the planar sensor is converted to the horizontal axis, and the data of elements 22-1, 23-1, 24-1 and elements 22-2, 23-2, 24-2 are made parallel to the x-axis. Although the data are lined up, the data is stair-like with a deviation of τ/3. Therefore, by performing this delay correction every τ/3, the entire data can be converted into linear data parallel to the x-axis.

Here, the relationship between delay times will be explained using the time chart shown in FIG. First, the time series follows the timing every τ/3 as in Fig. 5, and each number 0, 1, 2, 3... is t, t+τ/3, t+2/3 τ, t+3τ/3... It represents the time of the day. In addition, the imaged planar data is the elements 22-1, 23-
1, 24-1, 22-2, 23-2, 24-2,
As shown in 22-3, the vertical axis is read out first, and when the reading of this vertical axis is completed, the horizontal axis is read out. This imaging time advances every τ/3, and during this imaging time, the elements 22-1, 23-
The data of 1,...24-5 is being read. This figure shows the timing relationship centered around the imaging time "10". Delay amounts “6” (2τ) and “3” (τ) are determined by the gate of the demultiplexer 13, and delay amounts “0 to 4” (0, τ/3, 2τ/3) are determined by the gate of the synthesis/separation circuit 30. ,3τ/3,4τ/3
) is determined. The total delay amount in this case is "4, 7, 10, 3, 6, 9..." as shown in the figure, indicating the delay amount at arrow "A" in FIG. When this delay amount is subtracted from the imaging time, the output time is “6,
3, 0, 7, 4, 1...'', which means that the data captured at these times are output in that order. Note that this imaging time is 9, 10,
Although the reading during 11 is shown to be performed continuously, it is also possible to read at high speed at any time between these imaging times.

In the embodiment shown in FIG. 6b, the delay circuit is divided into two stages to make the explanation easier to understand, but the total delay amount is "4, 7, 10, 3,
6, 9, . . .'' to perform processing without dividing into two stages.

Generally, in a high-precision system using a multi-element image sensor, it is necessary to correct sensitivity non-uniformity of each element, so
Since a correction circuit is required, the addition of the correction circuit in FIG. 6 does not result in a large proportion.

As explained above, the present invention uses a two-dimensional sensor with a normal two-dimensional array, without developing an image sensor with a special array.
Since the resolution can be increased, a highly flexible and high-performance one-dimensional imaging device can be obtained.

The above explanation is based on the case where the angle θ is set to the value determined by equation (1) and sampling is performed at equal intervals on the x-axis. However, the angle θ may be set to be slightly different from this value. , the effect of increasing the number of samplings remains unchanged, and the value of the angle θ may be measured and corrected in advance. In addition, this example
Although the case where the light-receiving elements are arranged in a normal grid-like arrangement has been described, it is clear that the present invention can also be applied to a two-dimensional sensor having a checkered pattern or other special pattern shape.

Furthermore, the present invention can also perform multispectral observations. In other words, there is a considerable margin in the number of normal secondary elements in the vertical array, and when used in conjunction with multispectral observation, the number of elements in the two-dimensional array can be used effectively, as shown in Figures 8 and 9. You can also do it. These figures show the case where the number of observed spectra k = 4 and the sampling magnification n = 3. In Figure 8, the number of elements is increased by the number of spectra with the same angle setting as in Figure 4, so the tilt angle θ is the same. However, in FIG. 9, it can be considered that sensors of another spectral number (3 in this case) are inserted between these element intervals, so the inclination angle is as small as θ/4.

[Brief explanation of the drawing]

1 and 2 are block diagrams of the sensor section of a conventional high-resolution imaging device, FIGS. 3a and 3b are block diagrams of the transmitter and receiver of the high-resolution imaging device of FIG. 2, and FIG. A configuration diagram of a sensor unit according to an embodiment of the present invention, FIG. 5 is an explanatory diagram explaining the time relationship in FIG. 4, and FIGS. 6 a and b are transmissions of an embodiment of a high-resolution imaging device using FIG. 4. block diagram of the section and receiver section,
FIG. 7 is a time chart showing the time relationship in FIG. 6, and FIGS. 8 and 9 are configuration diagrams of second and third embodiments of the sensor portion of the present invention. In the figure, 1, 4, 5, 6...image sensor, 2, 3...
...Light receiving element (element) section, 7, 25...Amplifier, 8, 26...Sample and hold circuit, 9, 2
7... A/D converter, 10, 19... Multiplexer, 11... Modulation transmitter, 12... Reception demodulator, 13... Demultiplexer, 14, 15, 2
8, 29, 31-34...Delay circuit, 16, 1
7, 18... Correction circuit, 20... Image processing section, 2
1...Two-dimensional image sensor, 22-24-1,
2, 3... Light receiving element, 30... Combination separation circuit, 3
5...Synthesis circuit.

Claims (1)

[Claims] 1. An optical means for condensing a light image of an imaged object relatively moving in a predetermined direction, and a light-receiving surface in which a plurality of light-receiving elements are arranged in a lattice shape are tilted at a predetermined angle with respect to an imaging axis perpendicular to the direction of movement. an imaging means disposed in the condensing section of the optical means; a readout means for reading out the output of each light receiving element of the imaging means in accordance with a predetermined clock; and each of the plurality of predetermined rows of light receiving elements in the lattice arrangement. a delay amount corresponding to the time required for each of the plurality of predetermined rows of light receiving elements to move to a predetermined position parallel to the imaging axis so that the output signal becomes one-dimensional data parallel to the imaging axis; a high-resolution imaging device, the high-resolution imaging device comprising: delay correction means for applying the following to the output signal of each of the light-receiving elements.