JP2914054B2 - Imaging radiometer with band-to-band registration function - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an imaging radiometer, and more particularly to a multi-band imaging radiometer which is mounted on a mobile object such as an artificial satellite or an aircraft and which images a target area on the ground surface or the like in a band shape from the sky.

[0002]

2. Description of the Related Art An imaging radiometer of this type is used for remote sensing a target area along a trajectory of a moving object.
A multi-spectral scanner (MSS), which is a typical imaging radiometer mounted on Landsat which is one of the artificial satellites for remote sensing of (A), and a sematic mapper (TM) are well known. Also, as a recent one, for example, the Procedures of the Forth International Symposium on Space Technology and Science (PROCEEDINGSOF) published in Tokyo in 1984.
THE FOURTEENTH INTERRNATI
ONAL SYNPOSIUM ON SPACE A
ND SCIENCE), pages 1313 to 1319, "The Development of Multispectral Self-Scanning Radiometer for MOS-1" (THE DEVELOPMENT)
OF MULTI-SPECIAL ELECTRO
NICSELF-SCANNING RADIOMET
ER FOR MOS-1) and the 34th Congress of the International Astonautical Federation held in Budapest in 1983 (MESSR). 34th CONGRESS OF THE INT
ERNATIONAL FEDERATION), article IAF-83-109, "The SPOT-HRV.
Instrument: An Overview of Design and Performance "(The SPOT-H
RVInstrument: AnOverview
f Design and Performance)
HRV (High Resolution) described in
on Visible camera). The MESSR and HRV are linear array CCDs in which a plurality of photoelectric elements are linearly arranged as photoelectric detection elements.
This is an image-capturing radiometer using this type.

[0003] These imaging radiometers measure the intensity of visible light or infrared light radiated or reflected from the imaging region, ie, the target region, on the earth's surface, ie, radiance, and measure the brightness corresponding to the radiance. Is generated. Also, generally, these imaging radiometers
It has a so-called multi-spectral band (hereinafter abbreviated as multi-band) type configuration in which the visible light or infrared light is divided into a plurality of wavelength bands, that is, spectral bands, and an image is generated in each band. A camera section of the imaging radiometer for performing the imaging includes an optical system that focuses radiation / reflection from the target area into a radiation flux corresponding to the radiance, a photoelectric tube and a linear array CCD that convert the radiation flux into an electric signal. And a detector including a photoelectric element. The radiant flux is divided into the plurality of bands by a split optical system including a band separation mirror or a prism disposed in the optical path, and each band is detected by a dedicated band detector disposed at the focal position of each band. Converted to electrical signals. Linear array CC
The MESSR using D as a detector has three bands of visible light,
It has a total of four spectral bands of one infrared light band. Therefore, the number of divisions of the radiant flux is 4, and four detectors corresponding to each band are provided. If the positional relationship between the detectors in the plurality of bands or the optical path of the optical system corresponding to each detector does not exactly match the design value, the positional deviation between the pixels of the received image of each band is reduced. Occurs. The registration between the pixels of this image is called registration. In addition, the alignment between the multi-band images as described above is particularly called inter-band registration.

In an imaging radiometer of the mechanical scanning type using a vibrating or rotating mirror such as the above-mentioned MSS or TM, the number of photoelectric elements of the detector may be several or at most several for each band. Inter-band registration can be easily established. In the case of the MSS, six photoelectric elements are provided for each band, and in the case of the TM, sixteen photoelectric elements are provided for each band (only one band has four photoelectric elements). The photoelectric element pitch corresponding to the pixel pitch is also 100 μm. It is relatively large. Therefore, the alignment error of the photoelectric element allowed during the assembly of the detection unit reaches about one tenth of the pixel pitch, which is considerably large. Further, the fluctuation of the registration between the bands due to external factors such as the impact at launch and the temperature fluctuation is very small.

On the other hand, in a multi-band imaging radiometer such as the above-mentioned MESSR or HRV, which includes a linear array CCD as a radiation detector, a plurality of CCD detectors of each band-corresponding CCD detector are used to establish registration between bands.
It is necessary to perform the above alignment for each of the D elements.
More specifically, the linear array CCD, which is the detector of the MESSR, has C
Since the number of CD elements is 2048, the element pitch is 14 μm, and the focal plane length is 28.67 mm, if the allowable error of the inter-band registration is within 0.1 pixel similar to the above-mentioned MSS, the distance between the element positions of each band is different. The maximum permissible error in the lengthwise direction is 1.4 μm. CCD currently under development
In an electronic scanning type multi-band imaging radiometer with about 5000 elements, the element pitch is about 7 μm, so the maximum allowable error of the inter-band registration is 0.7 μm.
m. It is very difficult with the current state of the art to keep the error within this range. That is, the MESSR or H
The RV has an essential advantage of electronic scanning, that is, a large dwell time, that is, a time for gazing at a target in the case of the same resolution and the same scanning width, allows a reduction in the size of the optical system. Although it has the advantage of simplicity, it involves the problem that the inter-band registration is difficult.

[0006] Remote sensing by a multi-band imaging radiometer not only processes images of a plurality of bands individually, but also extracts features by comparing the features of the image of the observation target surface for each spectral band. In many cases, a multi-band image obtained by superimposing images of a plurality of bands is used. On the other hand, the above-mentioned conventional imaging radiometer does not have a function of detecting an abnormality in the adjustment state of the inter-band registration. Therefore, adjustment and calibration of the inter-band registration need to be performed at the time of manufacturing and at the time of mounting work on the moving body. Further, after the launch of the moving body, it has been impossible to detect the above-described band-to-band registration deviation due to a deformation of the structure or a deformation of the detector mounting portion due to an impact or a temperature change at the time of the launch. . For this reason, the registration between the bands after the launch has been nothing but so-called geometric correction based on the position of a specific target on an actual image in image processing on the ground. The method using the geometric correction selects the specific target on the image of a certain band, for example, a target having a characteristic and a large output level difference on the ground such as the tip of a peninsula, and determines the pixel position and time of the output image. As a reference, the image processing parameters are corrected by measuring the deviation of the position of the specific target on the image of another band in pixel units. According to this method, it is not possible to cope with an increase in the number of pixels and an increase in the amount of image processing accompanying an increase in resolution and an increase in scanning width. That is, the information to be processed for the geometric correction increases with the increase in the resolution and the increase in the scanning width, and the miniaturization of the image processing apparatus and the speeding up of the image processing are restricted. Furthermore, there is a problem that the selection of the specific target is restricted by the weather condition of the target area, and the geometric correction accuracy is reduced. Referring to FIG. 7 showing a remote sensing system using a conventional imaging radiometer, a conventional imaging radiometer 11 mounted on an artificial satellite 10 will be described.
Scans the ground surface in a band shape, for example, in the direction of arrow F with a scanning width of 100 km, and converts the data of the target area 30 to a ground resolution of 50 km.
Collect at m. The collected data is collected by the imaging radiometer 11
And is transmitted to the ground station 21 by the wireless data link transmitter. Also, from the ground station 21, an artificial satellite according to a preset schedule
Position information on the orbit of the satellite 10 and control information for setting the sensitivity of the imaging radiometer 11 based on the distinction between the land area and the sea area of the target area 30 at the orbital position are commanded through a wireless data link as artificial satellites. Supply 10 The artificial satellite 10 receives these position information and control information and uses them as control signals for the imaging radiometer 11. As described above, the artificial satellite 10 takes a sun-synchronous quasi-return orbit, and has an orbit altitude of 900 km, an orbit inclination of 99 °, an orbital period of 103 minutes, that is, a scanning speed of the ground surface of about 6.7 km / s. The data transmission of the wireless data link is limited to the time when the artificial satellite 10 is visible from the ground station 21 and is about 1 per orbit.
5 minutes. Referring to FIG. 8, the imaging radiometer 11 includes:
Camera unit 12 including optical system 121 and detection unit 122
A signal processing unit 13 for transmitting data from the camera unit 12 to the ground station 21; a wireless data link transmitter 14 for transmitting the data to the ground station 21; and a data link for receiving a command from the ground station 21. A receiver 15 and a deformatter 16 for separating the command into a control command for the signal processing unit 13 and other commands for the artificial satellite 10 are provided. The detection unit 122 of the camera unit 12 includes a linear array CCD 4 having 2000 CCD elements.
Are arranged so as to correspond to the four observation target bands 1 to 4, respectively, and a ground surface resolution of 50 m is provided. Here, for convenience of explanation, two CCD arrays corresponding to two of the bands will be described. Just to explain.

Referring also to FIG. 9, the detector 122 includes a beam splitter 12 composed of a spectral element such as a prism for separating incident light into two spectral bands 1 and 2.
3 and the light receiving elements of bands 1 and 2, each having 2000 elements, an element pitch of 14 μm, and a focal plane length of 28.7.
mm CCDs 124 and 125 are provided.

Referring again to FIG. 8, the signal processing unit 13
Are amplifiers 131A and 131B for amplifying output signals of the CCDs 124 and 125 with a gain set by a gain control signal from the de-formatter 16, and A / D converters for converting output signals of these amplifiers into 8-bit digital signals, respectively. 132A / 132B, and a formatter 133 for time-division multiplexing the digitized data and a telemeter signal for monitoring the operation state of the artificial satellite 10.

Referring to FIG. 10, a ground station 21 shown in FIG. 10 includes a console 211 for setting a control command including a sensitivity setting command, a transmission processor 212 for converting the control command into a signal for transmission, Machine 213
A duplexer 214 that superimposes and separates the transmission and reception, an antenna 215 that tracks the artificial satellite 10, a receiver 216, and a high-density recording device (HDDT) 220 that temporarily records data received by the receiver 216. ,
HDDT by referring to data from database 219
The image processing apparatus includes a data processing device 217 that processes data received from the CPU 220 and outputs image data, and an image output device 218 that converts the image data into an image and outputs the image data. The database 219 includes geometric correction data for an inter-band registration process.

The operation of this conventional example will be described with reference to FIGS.
1, based on the imaging schedule of the target area 30,
The information on the orbit of the artificial satellite 10 and the information indicating the distinction between the land area / sea area and the plain / mountain area of the target area 30 at the orbital position are generated.
Is set to 0 dB. In the transmission processor 212, the sensitivity setting command CS is multiplexed with other control commands of the satellite 10, and is transmitted as a control command to the transmitter 2.
13 is supplied. The transmitter 213 transmits the control command to the artificial satellite 10 via the duplexer 214 and the antenna 215. The data link receiver 15 of the artificial satellite 10 receives this and inputs it as a control signal CSA to the deformatter 16. Deformatter 16
Extracts a control command for the signal processing unit 13, decodes the sensitivity setting command CS among them, and supplies it to the amplifiers 131A / 131B as a sensitivity control signal SS.
The amplifier 131A / 131B is set to the gain H corresponding to the sensitivity 0 dB by the sensitivity signal SS designating the sensitivity 0 dB as described above. Next, the CCD 124/12 of the camera unit 12
5 from the signal processing unit 13
131A / 131 with gain set to 0 dB sensitivity
B respectively. The output of these amplifiers is A / D
A / D conversion is performed by the conversion units 132A / 132B, respectively, to become 8-bit digital data D1 / D2. Next, these digital data D1 / D2 are subjected to parallel / serial conversion in the formatter 133,
The data is formatted together with the housekeeping data and the like. The output of the formatter 133 is transmitted by the data link transmitter 14 to the ground station 21.

In the ground station 21, the received data supplied to the data receiver 216 via the antenna 215 and the duplexer 214 is recorded in the HDDT 220, and further supplied to the data processor 217 for processing. The data processing device 217 refers to the data in the database 219 to correct the scanning distortion of the received data, the variation in the sensitivity of the detector, and the like to generate image data. These processes are performed for individual observation bands 1 and 2, ie, CCDs 124 and 12
5 for each output data. The resulting image is thus a single band image of bands 1 and 2.
To generate a multiband image in which the single band images of these bands 1 and 2 are superimposed, the CCD 124/125 stored in the database 219 is further added.
A geometric correction process for inter-band registration is performed using the alignment data. The single band data output data of each of the bands 1 and 2 from the data processing device 217 is imaged as a single band image of each of the bands 1 and 2 by an image output device 218 using a high-density film recorder. The brightness of the pixels of this image is proportional to the level of the image data, and the image is monochrome. Any color for identification of imaging in the band data of each of the normal bands 1 and 2,
For example, red is assigned to band 1 and green is assigned to band 2 in advance. Thereby, the output image becomes a multi-color image composed of red and green. The multi-color image obtained by this method is called a pseudo-color image. The transmission of the command and the reception of the imaging data are performed during a transmission time of about 15 minutes of the wireless data link.

Next, referring to FIGS. 11A and 11B,
A method for adjusting and calibrating the inter-band registration on the ground will be described. Referring to FIG. 11A, a collimator 80 having a slit 81 attached to a focal point and a light source 82 is placed on an optical axis (Z
(Axis) are aligned. The longitudinal direction (X-axis) of the slit 81 is the CCD 124/125 of the radiometer 11.
Are set so as to be perpendicular to the arrangement direction (Y axis) of the photoelectric elements, that is, the width direction of the slit 81 is the Y axis. From the collimator 80, the collimated parallel light L having the shape of the slit 81 is incident on the radiometer 11, and the slit 81 is moved along the Y-axis.
Each of 125 photoelectric elements to be measured, for example, element outputs at the center and both ends (fifth, 1000th, and 1995th) are individually measured. The above movement is performed by the CCD 124
/ 125 is performed in a moving step corresponding to the element pitch of the photoelectric element. The above measurement result, CCD124 / 12
By comparing the positions of the slits 81 at which the element outputs of No. 5 have the highest levels, the deviation in element pitch units can be measured. Referring to FIG. 11B, the above measurement performed on the 1995th element shows that CCD1
It is shown that the shift of the elements between 24/125 is equivalent to two element pitches. Therefore, the CCD 124 /
The detector 125 adjusts the root mean square value of each of the deviations measured for the elements at the center and both ends so as to minimize the deviation, and incorporates the adjusted value into the detector 122. The final deviation data obtained as a result of this adjustment is the alignment data of the CCDs 124/125.

The focal length of the collimator 80 is set to be about 10 times the focal length of the radiometer 11 to be tested.
Double. The focal length of the radiometer 11 of this example is about 250 mm
Then, the focal length of the collimator 80 is 2.5 m, and the element pitch is 14 μm as described above, so that the moving step is 140 μm.

The CC of the radiometer 11 thus obtained is
The alignment data of D124 / 125 is
Is stored in a database 219 for geometric correction of image processing in the data processing device 217. The position data and attitude data of the artificial satellite 10 in orbit are compared with the read output of the database 219, and the CCD 124
The geometric correction is performed for each band 1/2 corresponding to the output of / 125. Thereby, registration calibration between bands on the output image is performed.

[0015]

The above-described conventional imaging radiometer does not have a function of detecting an abnormality in the adjustment state of the inter-band registration. It is not possible to detect the above-mentioned band-to-band registration deviation due to deformation of the structure or deformation of the detector mounting part due to changes in temperature or temperature, etc. In the processing, there is no other than geometric correction based on the position of a specific target on the actual image, but the geometric correction is performed by increasing the number of pixels and increasing the amount of image processing accompanying the increase in resolution and scanning width. However, there is a disadvantage that the information to be processed increases, and the miniaturization and speeding up of the image processing apparatus are restricted. Furthermore, there is a problem that the selection of the specific target is restricted by the weather condition of the target area, and the geometric correction accuracy is reduced.

Therefore, an object of the present invention is to provide an optical means for measuring the relative positional deviation of the detectors of a plurality of bands incorporated in a radiometer main body, and the inter-band registration based on the output of the means. And a means for performing a calibration function.

Another object of the present invention is to eliminate the need for processing of band-to-band registration by geometric correction in image processing on the ground, and thus to increase the speed of image processing and simplify the image processing apparatus. Is to provide a radiometer.

Still another object of the present invention is to provide an imaging radiometer which performs high-precision inter-band registration by the inter-band registration calibration function without being affected by the weather in the target area.

[0019]

SUMMARY OF THE INVENTION An imaging radiometer according to the present invention is mounted on a moving body that travels over an area to be imaged on the ground surface along an orbit having a predetermined positional relationship with the area. A unit imaging region defined by a predetermined width and a distance in the navigation direction of the moving object is included in the field of view, and visible light / infrared light from the unit imaging region is divided into first and second spectral bands. A first optical system for forming an image on each of the first and second focal planes, and a luminance distribution on the focal plane for the light beams of the first and second spectral bands respectively arranged on the focal plane. First and second photoelectric conversion means each comprising a plurality of photoelectric conversion elements arranged in an array so as to be detectable, wherein the unit imaging which enters the field of view as the moving object travels An imaging radiometer that sequentially captures optical images in the first and second spectral bands in a region, including a movable mirror that can be selectively inserted into an optical path of the first optical system. The optical image of the calibration pattern of the striped pattern for testing whether each of the photoelectric conversion elements of the photoelectric conversion means is at a position corresponding to each other on the first and second focal planes is formed by the first and second optical conversion means. A calibration pattern forming unit including a second optical system formed on the second photoelectric conversion element and an optical path switching unit for controlling insertion of the movable mirror into the optical path are provided.

[0020]

Next, an embodiment of the present invention will be described with reference to FIGS. 1 to 4, 5A and 5B. Referring to FIG. 1 which is a block diagram showing an embodiment of the present invention, an imaging radiometer 17 of this embodiment has an optical system 121 similar to that of the above-described imaging radiometer 11 and a detection with improved resolution described later. A camera unit 71 comprising a unit 172 and a calibration unit 180 for an inter-band registration test according to the present invention, a signal processing unit 13 comprising the same components as the conventional one, and the above-mentioned ground station 21
A data link transmitter 14 for transmitting imaging data to the data link receiver 15, a data link receiver 15 for receiving a command from the ground station 21, and a control command from the data link receiver 15 for transmitting a control command to the signal processing unit 13 and the camera unit 71. And a de-formatter 16 for separating the control commands from other control commands.

The bands to be observed and the scanning width of the imaging radiometer 17 are 4 bands and 100 km as in the conventional example, and the resolution is 20 m, which is much higher than the conventional example. The detection unit 172 of the camera unit 71 has a pitch of 7 μm so as to give sufficient ground resolution to each of the observation target bands 1 to 4.
And four linear array CCDs composed of 5000 CCD elements. Here, for convenience of explanation, only these four two parts will be described.

Referring also to FIG. 2, a detector 172 includes a beam splitter 173 similar to the detector 122 described above,
CCD 174/175. The calibration unit 180 includes a lamp 181/187 as a calibration light source and a lamp 181/18.
Diffusion plate 18 for equalizing the irradiation light from 7
2/188, a pattern 183 for calibrating the element array direction, and a pattern 18 for calibrating in the direction perpendicular to the element array direction.
9, a half mirror 184, and an optical system 185 for calibration
During observation, the incident light from the optical system 121 of the radiometer main body is retracted out of the optical path so as to be incident on the detection unit 172. During calibration, the incident light from the optical system 121 is cut off and the optical system 185 for calibration is used. Light from the detector 17
And a movable mirror 186 that makes the light incident on the second mirror 2.

At the time of normal observation, the ground station 21
The console 211 sets a command CU for instructing normal observation in addition to a control command including a sensitivity setting command CS for the artificial satellite 10, and transmits the command CU to the transmission processor 212, the transmitter 213, and the duplexer 21.
3. Transmit to the artificial satellite 10 via the antenna 215. This command CU is received by the data link receiver 15 and separated by the de-formatter 16. Based on the command CU, the movable mirror 186 of the calibration unit 180 is retracted to the observation position, that is, outside the main optical path, and the incident light from the optical system 121 is removed. The light enters the detection unit 173 and operates in the same manner as the above-described conventional example.

Next, when calibrating the inter-band registration, the ground station 21 sets a command CY for instructing execution of calibration in the element array direction (Y) from the console 211 and transmits the command CY to the artificial satellite 10 (step S1). ). Calibration unit 1 based on command CY from deformatter 16
The 80 movable mirror 186 enters the calibration position, that is, the main optical path, and the incident light from the optical system 185 enters the detection unit 173 (step S2). Next, the lamp 181 is turned on, the illuminance of the entire irradiation surface is made uniform by the diffusion plate 182, and the pattern 183 is irradiated.
5 to generate a projected image 183A on the light receiving surface of the CCD 174/175 by projecting the image on the detecting unit 172 (Step S).
3). Next, the CCD 174 /
The output signal A1 / A2 of 175 is amplified and A / D converted by the signal processing unit 13 and transmitted to the ground station 21 via the data link transmitter 14 as calibration data (step S).
4). Next, at the ground station 21, the projected image 183A is transmitted via the antenna 215, the duplexer 214, and the receiver 216.
Is recorded in the HDDT 220. This recording corresponds to one scene of the processed image, that is, about 70 km,
The scanning is performed for about 10.5 seconds from the above-described surface scanning speed of the artificial satellite 10 of 6.7 Km / S (step S5). Next, the ground station 21 uses the console 211 to set a command CX for instructing calibration execution in the direction (X) perpendicular to the element array, and transmits the command CX to the artificial satellite 10 (step S6). Based on the command CX from the deformatter 16, the lamp 181 is turned off, the lamp 187 is turned on, and the pattern 189 is irradiated via the diffusion plate 188.
85, the projected image 1 on the light receiving surface of the CCD 174/175
89A is generated (step S7). Next, the output of the CCD 174/175 based on the projected image 189A is used as calibration data via the data link transmitter 14 and transmitted to the ground station 2.
1 (step S8). Next, in the ground station 21, the data of the projection image 189A is similarly recorded in the HDDT 220 for about 10.5 seconds (step S9). When the acquisition of the series of calibration data DY / DX is completed, the ground station 21 sets the command CU again and transmits the command CU to the artificial satellite 10 (step S10). In the imaging type radiometer 17,
The movable mirror 186 of the calibration unit 180 based on the command CU
Is set to the observation position (step S11), and normal observation is executed (step S12).

In the image processing, referring to FIG. 5B showing the procedure for determining the pixel shift R between the bands in the element arrangement direction (Y) in the above-mentioned step 13, the calibration data recorded in the HDDT is stored in the CCDs 174/175, respectively. The image data is processed by the data processing device 217 for each output data. First, a pixel displacement R between bands in the element array direction is calculated by a procedure described in detail later (step S1).
3). Next, a pixel shift T between bands in a direction perpendicular to the element array direction is calculated (step S14).

The inter-band registration calibration of the present invention is based on an optical vernier scale determined by the element pitch of the CCD and the pitch of the grid pattern of the projected image. The details of this optical vernier will be further described.

Referring to FIGS. 3 and 4 showing an example in which the inter-band registration calibration is performed between the 4990th to 5000th light receiving elements of the CCDs 174 and 175, the CCDs 174/175 of the detection unit 172 are shown.
183 from the calibration unit 180 on the light receiving surface of
Is a grid-like pattern which is perpendicular to the element arrangement direction of the CCDs 174/175 and has a strip-shaped light portion and a shadow portion having the same width and slightly different from the element pitch. The output level of each element of these CCDs 174/175 is determined by the ratio of the area of the light part on the light receiving surface to the area of the mask part by the shadow part, and the sum of the light part and the shadow part of the projected image 183A. The output level periodically fluctuates due to the pitch P of the grating, which is the width of the element, the element pitch S of the CCDs 174/175, and the displacement between the CCDs 174/175 in the element arrangement direction.

The cycle L of the output level fluctuation determined by the grating pitch P is expressed by the following equation.

L = P · n (n is an integer) ····················· (1) Further, the period L has the following relationship with the element pitch S: .

L = S · m (m is an integer) ·········································································································· P · m S = 1 The minimum length L that satisfies (3) is L. For example, the element pitch S =
If 7 μm and the lattice pitch P = 12.5 μm, (3)
From the equation, n = 4 (m = 7), and therefore the period L
= 49 μm. That is, it is equivalent to seven element pitches.

Next, the output of the CCD is subjected to sensitivity correction and dark area output correction for each element, and then a theoretical value of an element number versus an output level is obtained for each of odd-numbered and even-numbered elements. The theoretical value is a straight line that monotonically increases or decreases in accordance with the element number. In FIG. 4, a solid line a and a dotted line b show examples of the straight lines corresponding to the even and odd numbers, respectively.

As is well known, the vernier scale is an auxiliary scale for further subdividing the scale and reading it, and has a scale obtained by equally dividing the length of the main scale u-1 or u + 1 scale by u. The length l of the vernier scale is expressed by the following equation, where s is the interval between the main scales and 1 is the interval between the subscales.

1 = upp = (ru ± 1) s Here, r is a small integer and is generally 1 or 2.

The relationship between the main scale and the vernier scale based on the grating pitch P and the element pitch S can be considered as a vernier scale including the light portion and the shadow portion of the grid with the element as the main scale. That is, the scale p of the vernier scale described above is １ ／ of the grid pitch P, that is, P / 2, the scale s of the main scale is the element pitch S, and u is 2n.
, And the cycle L of the output fluctuation corresponds to the length l of the vernier scale. However, since the shadow portion does not contribute to the photoelectric conversion output of the element, every other scale in the main scale / vernier relationship is effectively effective.

Next, two CCDs 1 having the same element pitch S
A case where 74/175 are shifted from each other by R in the element arrangement direction will be described. In FIG. 4, the CCD 174 /
An example in which the output levels of the 4990th to 5000th elements of each of the CCDs 175 are plotted is shown in FIG.
9 is indicated by a solid line M and a dotted line N. As described above, the output is an 8-bit digital signal whose level is in the range of 0 to 255 obtained by converting a binary number to a decimal number. Assuming that a shift in the element arrangement direction between the solid line M and the dotted line N at an arbitrary level 255 is ΔL, the shift R is expressed by the following equation (4) based on the theory of vernier scale using the element pitch as a unit.

R = (ΔL / L) · (P / 2S)... (4) If the deviation R is 0, ΔL = 0, that is, the solid line M and the dotted line N are shown as one overlapping line. ΔL takes two values depending on whether the element number is smaller or larger. In this example, since ΔL corresponds to a shift of 3 element pitches from the larger element number or 4 element pitches from the smaller element number, the shift R is 0.375 in the calculation from the former and 0.5 in the latter. Either. On the other hand, as described above,
The output level of each element is determined by the ratio of the area of the light part on the light receiving surface to the area of the mask part due to the shadow part. Therefore, if the output levels are compared among the elements having the same number, they are represented by the ratio. The difference is substantially proportional to the deviation R.
Here, 4991 is used to determine which of them
Focusing on the output of each element, if the output level 255 of the CCD 175 is 1, the output level of the CCD 174 is 191 which is about 3/4. In addition, 49 of CCD175
The output level of the 92nd element is 31 or about 0.12
Therefore, the deviation R is intermediate between the 0.25 element pitch and the 0.38 element pitch. Therefore, the deviation R is 0.3
75 can be determined. This shift R is a calibration value of the registration between bands.

Referring to FIG. 6 showing an example of inter-band registration calibration in a direction perpendicular to the element arrangement direction, a projected image 189A of the calibration pattern 189 is a CCD 174.
This is a slit-shaped pattern having an angle of θ with respect to the element arrangement direction of / 175. The width of the slit is the same as the element pitch S, that is, 7 μm, and neither the upper edge nor the lower edge of the slit masks the element, for example, 49 μm.
The output level of the 97th element is set to 500 as a reference value VC.
Assuming that the output level of the 0th element is V1 / V2, the deviation T between the CCDs 174 and 175 in the direction perpendicular to the element arrangement direction is expressed by the following equation.

T = (V 1 −V 2) / VC (5) The unit of the deviation T is the vertical size of the element. Also, C
The inclinations θ1 and θ2 of the CDs 174/175 with respect to the respective element arrangement directions are expressed by the following equations.

Θ1 = {1- (V1 / VC)} S / 4S (6) θ2 = {1- (V2 / VC)} S / 4S (7) If VC = 256, V1 = 241, and V2 = 231,
The deviation T is 0.039 element pitch, and the inclinations θ1 and θ2 are 14.6 and 24.6 (mrad), respectively.
That is, 50 minutes and 84 minutes.

Referring to FIG. 5B, first, in step S3
With respect to each of the calibration data DY1 / DY2 of the CCDs 174/175 collected in steps S5 to S5, pixels are allocated to odd-numbered and even-numbered pixels, respectively, and are arranged as imaging data. That is, the odd numbers are arranged in the order of 1,3,5... 4997, 4999th, and the even numbers are arranged in the order of 2,4,6... 4998,5000th (step S131). The odd-numbered data strings are OY11 and OY21, and the even-numbered data strings are EY11 and EY21. Next, the data level of each pixel of OY21 is inverted to generate an inverted OY21 (step S132). Next, OY11 and inverted OY21
The data of the inversion OY21 is shifted in the pixel array direction so that the levels of all the pixels become minimum when the data of the corresponding pixels are added. In this embodiment, the above condition is satisfied when the inverted OY21 is shifted to the left by three pixel pitches for the 4990th to 5000th data.
This shift amount is the shift ΔLO between the odd-numbered pixels (steps S133 to S136). Similarly, the shift ΔLE between the even-numbered pixels is obtained (step S137).
Next, an average value of these ΔLO and ΔLE, that is, a deviation ΔL between pixels is obtained (step S138). When the above-described added data is imaged for both the odd-numbered and even-numbered data strings, a substantially uniform image can be obtained on the entire screen at a low level. In this manner, the difference portion having a value different from ΔL becomes extremely bright and can be easily identified.
On the other hand, the period L is a constant set from the grating pitch P of the projected image 183A determined by the grating pitch of the calibration pattern 183 and the magnification of the optical system 185, and the element pitch S of the CCD 174/175. The calibration value R of the translation can be easily obtained by the equation (4) (step S139).

Next, the procedure for obtaining the pixel shift T between the bands in the direction (X) perpendicular to the element arrangement direction in step 14 is as follows. First, the CCD 1 collected in steps S6 to S8
Each of the calibration data DX1 / DX2 of 74/175 is imaged. The element number to be measured, in this embodiment, the strongest level pixel near the 5000th, ie, the 4997th pixel, is extracted, and this output level is used as the measurement reference value VC. The level V1 / V2 of the 5000th pixel of the calibration data DX1 / DX2 is measured, and the shift T in the direction perpendicular to the element array direction is obtained by the equation (5). The measurement of these levels is well known in image processing, and a description thereof will be omitted. The geometric correction parameters of the image processing stored in the database 219 are corrected using the calibration values R and T of the inter-band registration (step 15).

[0042]

As described above, the imaging type radiometer according to the present invention includes a movable mirror which can be selectively inserted into the optical path of the first optical system, and each of the first and second photoelectric conversion means is provided with a movable mirror. An optical image of a calibration pattern of a striped pattern for testing whether the photoelectric conversion elements are at positions corresponding to each other on the first and second focal planes is formed on the first and second photoelectric conversion elements. By providing a calibration pattern forming means including a second optical system and an optical path switching means for controlling insertion of the movable mirror into the optical path, an inter-band registration is measured on the moving body, and the data thereof is measured. Can be used for image processing on the ground side,
There is no need for processing for band-to-band registration, so that there is an effect that image processing can be speeded up and image processing can be simplified.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of an imaging radiometer according to the present invention.

FIG. 2 is a diagram schematically showing a configuration of an optical system included in an embodiment of the present invention.

FIG. 3 is a diagram showing a state in which a registration calibration light pattern is projected on a light receiving element of each band.

4 is a diagram showing output levels of light receiving elements of each band in the state of FIG.

5A and 5B are a flowchart of calibration of inter-band registration and a flowchart of calculation of inter-band pixel shift in the embodiment.

FIG. 6 is a diagram showing a state in which a slit-shaped light pattern having an angle of θ with respect to the pixel array is projected on a light receiving element.

FIG. 7 is a diagram of an example of a remote sensing system using a conventional imaging radiometer.

FIG. 8 is a block diagram of an example of a conventional imaging radiometer.

FIG. 9 is a configuration diagram of an optical system of the imaging radiometer of FIG. 8;

FIG. 10 is a block diagram illustrating an example of a configuration of a ground station in the system of FIG. 7;

FIG. 11 is an explanatory diagram of a method of adjusting and testing a band-to-band registration of a conventional imaging radiometer, and a graph of an example of a test result.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Artificial satellite 11, 17 Imaging radiometer 12, 71 Camera part 13 Signal processing part 14 Data link transmitter 15 Data link receiver 16 Deformatter 21 Ground station 30 Target area 80 Collimator 81 Slit 82 Light source 121, 176, 185 Optical system 122, 172 Detector 123, 173 Beam splitter 124, 125, 174, 175 CCD 131A, 131B Amplifier 132A, 132B A / D converter 133 Formatter 180 Calibrator 181, 187 Lamp 182, 188 Diffusion plate 183, 189 Pattern 183A, 183B Projection image 184 Half mirror 186 Moving mirror 211 Console 212 Transmit processor 213 Transmitter 214 Duplexer 215 Antenna 216 Receiver 217 Data processing device 218 Image output device 219 Database 220 HDDT

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ identification code FI G01J 5/48 G01J 5/48 EF (58) Field surveyed (Int.Cl. ⁶ , DB name) G01J 1/02 B64D 47 / 08 B64G 1/66 G01J 1/44 G01J 5/48

Claims (9)

(57) [Claims] 1. A predetermined width of an area to be imaged and a distance in a navigation direction of the moving object mounted on a moving object navigating along a trajectory having a predetermined positional relationship with the area to be imaged on the ground surface. And a visible / infrared ray from the unit imaging region is divided into first and second spectral bands and focused on the first and second focal planes, respectively. A first optical system to be imaged, and a plurality of optical systems each arranged in an array so as to be able to detect the distribution of the luminance of the light beams of the first and second spectral bands on the focal plane, respectively arranged on the focal plane. And first and second photoelectric conversion means comprising a photoelectric conversion element, wherein the first and second spectral bars of the unit imaging area entering the field of view as the moving object travels. An imaging radiometer that sequentially captures an optical image of the first and second photoelectric conversion units, wherein the imaging type radiometer includes a movable mirror that can be selectively inserted into an optical path of the first optical system. Forming an optical image of a striped calibration pattern on the first and second photoelectric conversion elements for testing whether the elements are at positions corresponding to each other on the first and second focal planes; 2. An imaging radiometer comprising: a calibration pattern forming unit including the optical system of No. 2; and an optical path switching unit that controls insertion of the movable mirror into the optical path. 2. The method according to claim 1, wherein the optical image of the calibration pattern is the first image.
And a width substantially the same as the light receiving surface of the second photoelectric conversion means, a length substantially equal to the photoelectric conversion element, and a pitch of the stripe pattern being slightly different from an arrangement pitch of the photoelectric conversion elements. The imaging radiometer according to claim 1, wherein the pattern is a calibration optical image pattern. 3. An optical image of the calibration pattern is a pattern of a second calibration optical image of a stripe pattern extending at a predetermined angle with respect to the arrangement direction of the photoelectric conversion elements. Item 9. An imaging radiometer according to Item 1. 4. The optical path switching means retracts the movable mirror from the optical path of the first optical system to outside the optical path when imaging the unit imaging area, and moves the movable mirror to the optical path when the calibration pattern is generated. 2. An imaging radiometer according to claim 1, including means for inserting. 5. A predetermined width of the area mounted on a moving body navigating along a trajectory having a predetermined positional relationship with the area to be imaged on the ground surface and a unit of the moving direction of the moving body. Having a unit imaging region elongated in the direction of the width defined by the distance in the field of view and dividing a light beam such as a visible ray from the unit imaging region into light beams of first and second spectral bands, and An optical system for forming an elongated first and second optical image on a second focal plane, respectively; and an optical system arranged on the focal plane so that the first and second elongated optical images can be respectively converted into electric signals. First and second photoelectric conversion means each comprising a plurality of photoelectric conversion elements linearly arranged at a predetermined element pitch along the optical image, and unit An imaging radiometer that sequentially captures the optical images in the first and second spectral bands of an imaging region as a chain of the optical images corresponding to the unit imaging regions, wherein the first and second photoelectric conversion units are provided. The optical images of the repeated dark stripe pattern having a pattern pitch and a width having a predetermined relationship with the element pitch so that a deviation from a position corresponding to each other in the first and second focal planes can be detected. Means for selectively forming at the position of the second photoelectric conversion means; and means for calculating the shift in response to the output of the first and second photoelectric conversion means at the time of forming the stripe pattern optical image. An imaging radiometer characterized by the above-mentioned. 6. A method for increasing the accuracy of calculating the deviation,
The imaging radiometer according to claim 5, wherein the element pitch forming at least a part of the photoelectric conversion means has a main scale / a vernier relation with the pattern pitch. 7. The shift calculating means compares the output of every other one of the plurality of photoelectric conversion elements of the first photoelectric conversion means with the output of a corresponding photoelectric conversion element of the second photoelectric detection means. 6. An imaging radiometer according to claim 5, including means for performing. 8. A slit orientation having a predetermined relationship with the element orientation so that deviations of the first and second photoelectric conversion means from the mutually corresponding element orientations at the first and second focal planes can be detected. 6. The imaging radiometer according to claim 5, further comprising: means for selectively forming an optical image of a slit pattern having a width and a width at the positions of the first and second photoelectric conversion means. 9. The imaging radiometer according to claim 5, wherein said stripe pattern optical image forming means, shift calculating means, and slit pattern optical image forming means are driven by a command signal from a ground station.