`U.S. Pat. No. 7,477,284
`IPR2013‐00327
`EXHIBIT
`Sony‐
`
`
`
`(19) Japanese Patent Office (JP)
`(12) Kokai Unexamined Patent Application Bulletin (A)
`(11) Laid Open Patent Application No.
`
`8-159762
`(43) Publication Date
`
`June 21, 1996
`Number of Claims
`
`25 OL
`
`Number of Pages
`
`15
`
`Examination Request
`
`not yet made
`
`
`
`Internal File No. FI
`
`
`
`9365-5H
`
`HV
`
`(51)
`
`
`Int. Cl.6
`G01C 11/06
`G01B 11/00
`G01C 3/06
`
`Identification Code
`
`
`(21) Application No.:
`
`6-298224
`
`
`
`(71)
`
`Applicant:
`
`(22) Application Date:
`
`December 1, 1994
`
`(72)
`
`Inventor:
`
`
`
`G06F 15/62 350A
` 415
`
`Tech. Indic.
`
`
`
`Continued on the
`last page
`
`000213909
`Aero Asahi Corporation
`3-1-1 Higashiikebukuro,
`Toshima-ku, Tokyo-to
`INOUE, Toru
`Aero Asahi Corporation
`3-1-1 Higashiikebukuro,
`Toshima-ku, Tokyo-to
`Patent Attorney, TANAKA, Tsuneo
`
`
`
`(74)
`
`
`
`
`Agent:
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`(54) [Title of invention] Three-Dimensional Data Extraction Method and Device, and Stereo Image Forming Device
`
`
`(57) [Abstract]
`[Object] To create DEM data from video images.
`[Constitution] Video images are taken of a target region
`from the air (S1). At this time, the camera position is
`measured by way of differential GPS. The camera is
`mounted on an anti-vibration device, and the orientation
`of
`the camera
`is measured precisely by way of
`gyroscope output and magnetic bearing sensor output
`thereof. Exterior orientation elements are determined
`accurately by matching fields in video images that
`overlap 60% (S2). The leading line, middle line, and final
`line of each field are extracted, and are separately
`combined
`to create continuous mosaic
`images
`consisting of a forward view image, a nadir view image
`and a rearward view image (S3). The vertical parallax is
`removed from the continuous mosaic images (S4). The
`parallax difference is calculated from the forward view
`image and the rearward view image (or the nadir view
`image) (S5), and the height is calculated from the
`parallax difference (S6).
`
`
`
`
`
`
`
`
`Translation by Patent Translations Inc. 1-800-844-0494 mail@PatentTranslations.com
`
`
`
`[Claims]
`[Claim 1] A
`three-dimensional data extraction method
`comprising: a basic information collection step of imaging a
`three-dimensional data extraction target while moving, recording
`that image signal, and recording imaging information including
`the position and orientation of the imaging camera; and a
`three-dimensional
`data
`generation
`step
`of
`generating
`three-dimensional data for said extracted object from the images
`and imaging information collected in said basic information
`collection step,
`being
`extraction method
`data
`the
`three-dimensional
`characterized in that said three-dimensional data generation step
`comprises:
`a continuous mosaic image generation step of extracting image
`data of prescribed lines in prescribed screens in consecutive
`screens of captured images, and generating at least two
`continuous mosaic images from among a forward view image, a
`nadir view image, and a rearward view image;
`a vertical parallax removal step of removing vertical parallax from
`the continuous mosaic images generated in said continuous
`mosaic image generation step,
`a parallax difference calculation step of calculating the parallax
`difference for a prescribed position in the continuous mosaic
`images from which vertical parallax was removed in said vertical
`parallax removal step; and
`a height calculation step of calculating the height of said
`prescribed position from the parallax difference calculated in said
`parallax difference calculation step.
`[Claim 2] The three-dimensional data extraction method recited
`in claim 1, further comprising an orientation calculation step of
`establishing, by way of relative orientation and successive
`orientation, exterior orientation elements
`from consecutive
`screens of captured images.
`[Claim 3] The three-dimensional data extraction method recited
`in claim 2, wherein said orientation calculation step comprises: a
`relative orientation step of extracting, from consecutive screens
`of captured images, two screens that overlap in a prescribed
`proportion and performing relative orientation; and a successive
`orientation step of associating models that have been relatively
`oriented by said relative orientation step.
`[Claim 4] The three-dimensional data extraction method recited
`in claim 3, wherein said prescribed proportion is 60%.
`[Claim 5] The three-dimensional data extraction method recited
`in any one of claims 2 to 4, wherein said vertical parallax removal
`step comprises: an exterior orientation element interpolation step
`of interpolating exterior orientation elements for each line of the
`continuous mosaic images generated in said continuous mosaic
`image generation step, in accordance with the exterior orientation
`elements determined in said orientation calculation step; and a
`projection step of transforming the lines of the continuous mosaic
`images generated in said continuous mosaic image generation
`step into images projected to a prescribed altitude in accordance
`with the exterior orientation elements of said lines.
`[Claim 6] The three-dimensional data extraction method recited
`in any one of claims 1 to 5, wherein said parallax difference
`calculation step comprises: an intermediate image formation step
`of forming one or more intermediate images between said
`continuous mosaic images; a corresponding point detection step
`of going through said one or more intermediate images and
`detecting corresponding points
`in said continuous mosaic
`images; and a computation step of calculating the parallax
`difference of said corresponding points in accordance with the
`detection results of said corresponding point detection step.
`[Claim 7] The three-dimensional data extraction method recited
`in claim 6, wherein said intermediate image formation step
`
`
`
`JP-08-159762-A Page 2
`extracts image data of intermediate lines in prescribed screens in
`consecutive screens of captured
`images, and
`forms said
`intermediate image.
`[Claim 8] The three-dimensional data extraction method recited
`in any one of claims 1 to 7, wherein a GPS reception means is
`provided as a means for detecting the position of said camera.
`[Claim 9] The three-dimensional data extraction method recited
`claim 8, wherein a correction means that differentially corrects
`the output of the GPS reception means is further provided.
`[Claim 10] The three-dimensional data extraction method recited
`in any one of claims 1 to 9, wherein said camera is prevented
`from vibrating by an anti-vibration means.
`[Claim 11] A
`three-dimensional data extraction device
`characterized by comprising: a video camera that images a
`three-dimensional data extraction target; a position measurement
`means that measures the position of said video camera; a
`recording means that records images captured by said video
`camera and measurement values of said position measurement
`means; a conveyance means that conveys said video camera
`and said position measurement system [sic]; a reproduction
`means that reproduces the image information and position
`information recorded by said recording means; a continuous
`mosaic image generation means that extracts image data of
`prescribed lines in prescribed screens in consecutive screens of
`captured images, and generates at least two continuous mosaic
`images from among a forward view image, a nadir view image,
`and a rearward view image; a vertical parallax removal means
`that removes vertical parallax from the continuous mosaic images
`generated by said continuous mosaic image generation means; a
`parallax difference calculation means that calculates parallax
`difference in continuous mosaic images from which vertical
`parallax has been removed by said vertical parallax removal
`means; and a height calculation means that calculates the height
`of said [sic] prescribed position from the parallax difference
`calculated by said parallax difference calculation means.
`[Claim 12] The three-dimensional data extraction device recited
`in claim 11, further comprising an orientation calculation means
`that establishes, by way of relative orientation and successive
`orientation, exterior orientation elements
`from consecutive
`screens of captured images.
`[Claim 13] The three-dimensional data extraction device recited
`in claim 12, wherein said orientation calculation means provides:
`a relative orientation step of extracting, from consecutive screens
`of captured images, two screens that overlap in a prescribed
`proportion and performing relative orientation; and a successive
`orientation step of associating models that have been relatively
`oriented by said relative orientation step.
`[Claim 14] The three-dimensional data extraction device recited
`in claim 13, wherein said prescribed proportion is 60%.
`[Claim 15] The three-dimensional data extraction device recited
`in any one of claims 12 to 14, wherein said vertical parallax
`removal means comprises an exterior orientation element
`interpolation means that interpolates exterior orientation elements
`in each line of the continuous mosaic images generated by said
`continuous mosaic image generation means, in accordance with
`the exterior orientation elements determined by said orientation
`calculation means; and a projection means that transforms the
`lines of the continuous mosaic images generated by said
`continuous mosaic
`image generation means
`into
`images
`projected to a prescribed altitude in accordance with the exterior
`
`
`Translation by Patent Translations Inc. 1-800-844-0494 mail@PatentTranslations.com
`
`
`
`
`
`JP-08-159762-A Page 3
`expense. If three-dimensional measurement is done by stereo
`matching using aerial photographs taken from a low altitude,
`matching errors will occur due to the influence of occlusion. This
`is because the two images that form a stereo image are seen
`from different viewing directions, and differences in the image
`due to differences in the observation direction make perfect
`matching impossible. In conventional cases, attempts have been
`made to eliminate such influence by using multiple orientation
`points on the ground, but with this, automation is impossible.
`[0003]
`[Problems to Be Solved by the Invention] In contrast, it is easy
`to automate the technology using video images to create
`topographical maps. But
`in
`the prior art,
`in extracting
`photographic three-dimensional data, in the same way as in an
`aerial survey, it is considered necessary to have several air-photo
`signal points
`in
`the
`image
`(signals
`for which clear
`three-dimensional coordinates are known), and even if the
`necessary number of air-photo signals are assured, the errors
`will be of poor precision, which is in meter units, making this
`impractical.
`[0004] Three-dimensional topographical maps are useful for the
`management of roads, rivers, railroads, and the like, and for
`planning new routes for the same and surveying the state of
`development of cities and the like; there is a demand for a
`system that makes it possible to acquire three-dimensional
`topographical
`data
`quickly,
`cheaply,
`and
`simply.
`If
`three-dimensional topographical data is available, bird’s-eye
`views can also be created easily (shown on a display or output
`on a printer), and various simulations can be run. And if
`three-dimensional data can be obtained by processing video
`images, it will also be easy to extract just the parts that have
`changed, thus also making it easy to survey the state of
`development of cities and the like.
`[0005] An object of the present invention is to propose a
`three-dimensional data extraction method and device, which
`extract three-dimensional data automatically.
`[0006] Another object of the present invention is to propose a
`three-dimensional data extraction method and device, which
`extract three-dimensional data from video images.
`[0007] Another object of the present invention is to propose a
`stereo image formation device that forms stereo images (two
`images suitable for stereo matching) from video images.
`[0008]
`[Means for Solving the Problems] In the present invention,
`image data of prescribed lines of video images are extracted, and
`continuous mosaic images having different parallax are formed.
`After vertical parallax is removed from these continuous mosaic
`images, the parallax difference is calculated by stereo matching.
`Heights are calculated from the resulting parallax differences.
`[0009] Preferably, at least three images that overlap in a
`
`orientation elements of said lines.
`[Claim 16] The three-dimensional data extraction device recited
`in any one of claims 11 to 15, wherein said parallax difference
`calculation means comprises: an intermediate image formation
`means that forms one or more intermediate images between said
`continuous mosaic images; a corresponding point detection
`means that goes through said one or more intermediate images
`and detects corresponding points in said continuous mosaic
`images; and a computation means that calculates the parallax
`difference of said corresponding points in accordance with the
`detection results of said corresponding point detection means.
`[Claim 17] The three-dimensional data extraction device recited
`in claim 16, wherein said intermediate image formation means
`extracts image data of intermediate lines in prescribed screens in
`consecutive screens of captured
`images, and
`forms said
`intermediate image.
`[Claim 18] The three-dimensional data extraction device recited
`in any one of claims 11 to 17, wherein said position measurement
`means is a GPS reception means.
`[Claim 19] The three-dimensional data extraction device recited
`in claim 18,
`further comprising a correction means
`that
`differentially corrects the output of the GPS reception means.
`[Claim 20] The three-dimensional data extraction device recited
`in any one of claims 11 to 19, wherein said camera is mounted on
`said conveyance means via an anti-vibration means.
`[Claim 21] The three-dimensional data extraction device recited
`in any one of claims 11 to 20, wherein said conveyance means is
`an aircraft.
`[Claim 22] The three-dimensional data extraction device recited
`in any one of claims 11 to 21, further comprising a bearing
`detection means that detects the bearing of said camera, wherein
`the output of said bearing detection means is also recorded in
`said recording means.
`[Claim 23] The three-dimensional data extraction device recited
`in any one of claims 11 to 22, wherein said camera is disposed
`so that the direction of travel of said conveyance means is a
`direction that is perpendicular to the direction of the scan lines of
`the camera.
`[Claim 24] A stereo image formation device characterized by
`comprising: an extraction means that extracts line image data, at
`two or more different prescribed line positions in screens, in a
`video captured image; and a combining means that combines
`line image data from the same line positions.
`[Claim 25] The stereo image formation device recited in claim 24,
`further comprising a vertical parallax removal means that
`removes vertical parallax from the images combined by said
`combining means, based on exterior orientation elements of each
`of the line image data on which [the images] are based.
`[Detailed Description of the Invention]
`[0001]
`[Field of Industrial Application] The present invention relates to
`a method and a device for three-dimensional data extraction, and
`to a stereo image formation device; more specifically, it relates to
`a method and device for extracting three-dimensional data from
`video images, and to a stereo image formation device that forms
`stereo images from video images.
`[0002]
`[Prior Art] Conventionally, aerial survey technology based on
`aerial photographs has been used in creating three-dimensional
`topographical maps. But aerial survey technology involves
`having a helicopter or light airplane fly above a locality while
`taking stereoscopic photographs of the ground, and analyzing the
`resulting
`stereo
`photographs;
`obtaining
`stereographic
`photographs alone requires a great deal of time and expense,
`and analysis of the same likewise entails enormous effort and
`
`
`Translation by Patent Translations Inc. 1-800-844-0494 mail@PatentTranslations.com
`
`
`
`
`
`JP-08-159762-A Page 4
`anti-vibration stabilizing device 12 can be set arbitrarily; the
`camera control device 22 controls the focus, zoom, diaphragm
`value, color balance and the like of the camera 10, and the VTR
`control device 24 controls the recording start, stop, and pause on
`the VTR 14, and also acquires and transfers to the computer 18
`a time code that is recorded together with the output image signal
`of the camera 10. This time code is used for synchronization
`when analyzing the image information and other measurement
`data that is recorded on the VTR 14, in the on-the-ground
`analysis system.
`[0017] A height above ground sensor 26 detects the height
`above ground, and the magnetic bearing sensor 28 detects the
`magnetic bearing. Because, even with
`the high-precision
`anti-vibration stabilizing device 12, there is slow directional
`movement caused by gyroscopic drift, it is necessary to correct
`the orientation of the camera by way of the magnetic bearing
`sensor 28. The outputs of the sensors 26 and 28 are applied to
`the computer 18 as digital data. Also input to the computer 18 is
`tri-axial gyroscopic data from the anti-vibration stabilizing device
`12 indicating the tri-axial orientation of the camera 10 (roll angle,
`pitch angle, and yaw angle), and zoom data indicating the zoom
`value from the camera 10.
`[0018] [Reference numeral] 30 is a GPS (global positioning
`system) receiving antenna, and 32 is a GPS reception device
`that collects the current ground coordinates (longitude, latitude,
`and altitude) from the GPS antenna 30. The GPS position
`measurement data output from the GPS reception device 32 is
`applied to the computer 18 for recording, and is also applied to a
`navigation system 34 for navigation. The navigation system 34
`performs
`three-dimensional graphic display of
`the current
`position on a screen of a monitor 38, with respect to set survey
`lines in accordance with navigational (survey line data) that has
`been previously recorded on a floppy disk 36. This makes it
`possible to capture images following along desired survey lines in
`regions where there are no target objects on the ground, or in
`regions where
`they cannot be ascertained (for example,
`mountainous regions or sea regions or the like).
`[0019] Note that, the differential GPS (D-GPS) method in which,
`even at reference points where the coordinates are known,
`measurements are taken by GPS, and the GPS position
`measurement data is corrected by the measurement error, is
`known as a way to improve GPS measurement precision. In this
`working example, this differential GPS method is adopted; the
`coordinates of a reference station of known coordinates are
`measured at the same time by GPS, and the measurement error
`data is radioed to the helicopter as GPS correction data. The
`communication device 40 receives the GPS correction data from
`the reference station and transfers it to the computer 18.
`[0020] The computer 18 records the flight data that is input
`(height above ground data, magnetic bearing data, zoom data,
`tri-axial gyroscopic data), together with GPS correction data and
`
`prescribed proportion are extracted from consecutive screens of
`captured images, the screens are matched by way of the
`overlapping parts thereof, and exterior orientation elements are
`established. Then, in accordance with the exterior orientation
`elements determined by this orientation calculation, the exterior
`orientation elements are interpolated for each line of the
`continuous mosaic images, and the lines of the continuous
`mosaic images are transformed into images projected to a
`prescribed altitude, in accordance with the exterior orientation
`elements of said lines.
`[0010]
`[Operation] The above processing can be automated on a
`computer, thus making it possible to automatically execute on a
`computer all the processes by which the stereo images needed
`for stereo matching are obtained from images resulting from
`video imaging, and by which heights are calculated, making to
`possible to quickly obtain three-dimensional data for a desired
`region or the like.
`[0011] Being video images, a great deal of information is
`available that is needed for establishing exterior orientation
`elements, and the precision of the exterior orientation elements is
`increased. Consequently, the height data that is ultimately
`obtained is also of good precision. Furthermore, in the stereo
`image matching computation as well, by temporarily creating
`intermediate images for said stereo images, and searching for
`corresponding points chainwise,
`the corresponding points
`between stereo images can be established with greater precision
`than in the case of stereo photographs, and the occlusion
`problem can be completely solved.
`[0012]
`[Working Example] Hereafter, referring to the drawings, a
`working example of the present invention is described in detail.
`[0013] FIG. 1 shows a schematic block diagram of an airborne
`measurement system in a working example of the present
`invention; FIG. 2 shows a schematic block diagram of an
`on-the-ground measurement system; and FIG. 3 shows
`schematic block diagram of an on-the-ground analysis system.
`[0014] The airborne measurement system shown in FIG. 1 will be
`described. In this working example, the airborne measurement
`system shown in FIG. 1 is aboard a helicopter. In this working
`example, a high-quality camera 10
`is mounted on a
`high-precision anti-vibration stabilizing device (anti-vibration
`device) 12, and the high-quality image signal output thereof is
`recorded on videotape by a high-quality video tape recorder 14.
`Note that, the camera 10 is generally facing downward, and is set
`so that the image directly below moves in a direction that is
`perpendicular to the scan lines. The output image signal of the
`camera 10 is also applied to a high-quality monitor 16. This
`allows visual confirmation of what the camera 10 is capturing and
`the state of imaging.
`[0015] The high-precision anti-vibration stabilizing device 12 is
`made so that vibration from the aircraft does not affect the
`camera 10. This makes it possible to record images without
`blurring. That is to say, by combining a gyroscope and gimbal
`servo, the high-precision anti-vibration stabilizing device 12 has a
`spatial stabilization function that keeps the optical axis of the
`camera 10 pointed in a fixed direction in inertial space against
`any fluctuations in the angles about the roll, pitch, and yaw axes
`that arise in the airframe.
`[0016] [Reference numeral] 18 is a personal computer that, along
`with collecting and recording measurement data, controls the
`high-precision anti-vibration stabilizing device 12 via a three-axis
`control device 20, controls the camera 10 via a camera control
`device 22, and controls the VTR 14 via a VTR control device 24.
`With the three-axis control device 20, the target bearing of the
`
`
`Translation by Patent Translations Inc. 1-800-844-0494 mail@PatentTranslations.com
`
`
`
`
`
`JP-08-159762-A Page 5
`recorded on a floppy disk 42 along with the time codes from the
`VTR 14. The time codes are used during analysis on the ground
`for synchronizing the position and orientation of the camera with
`the reproduced images.
`[0025] The position of the camera is known basically from the
`GPS position measurement data that is output from the GPS
`receiver [sic] 32, and for sake of better precision undergoes
`differential processing with the GPS correction data from the
`reference station. The differential processing may be done
`aboard the aircraft, but in consideration of such factors as bad
`communication of the GPS correction data, it is preferable to
`record the output of the GPS receiver 32 (the GPS position
`measurement data) and the GPS correction data separately on a
`floppy disk 42, and to do the differential processing during
`analysis on
`the ground. When
`there has been bad
`communication of the GPS correction data, then the GPS
`position measurement data undergoes differential processing
`with the GPS correction data that has been recorded and saved
`by way of the on-the-ground measurement system shown in FIG.
`2.
`[0026] With regard to the orientation of the camera 10, values
`wherein the output of the gyro sensors of the anti-vibration
`stabilizing device 12 have been corrected by the output of the
`magnetic bearing sensor 28 are recorded to the floppy disk 42.
`Specifically, the orientation of the three axes (pitch, roll, and yaw)
`is recorded on the floppy disk. Of course, the orientation of the
`camera 10 may be fixed for sake of simplicity, or if the
`performance of the anti-vibration stabilizing device 12 is good, or
`if simplification is acceptable.
`[0027] In this connection, the imaging altitude is set to 1000 feet,
`and with a Hi-Vision camera that uses a 2-million-pixel CCD
`image sensor, if the focal length is 8.5 mm, the imaging range is
`339 m and the resolution is 17.7 cm, and if the focal length is
`102.0 mm, the imaging range is 28 m and the resolution is
`1.5 cm.
`[0028] The recorded information (image and flight information) is
`reproduced and analyzed by the on-the-ground analysis system
`shown in FIG. 3. As described above, in accordance with the
`information during imaging from the computer 80 (the camera
`position and bearing, and the time codes), the workstation 76
`controls the VTR 70 and causes it to play back images with the
`same time code. The reproduced image signals are digitized and
`stored in the frame buffer 74. In this way, the workstation 76 can
`obtain image data as well as data on the position and orientation
`of the camera during imaging, and outputs DEM data after going
`through the various processing of orientation calculation (S2),
`continuous mosaic image formation (S3), vertical parallax
`
`timer codes from the VTR control device 24, to the floppy disk 42.
`The computer 18 can also display various input data on the
`monitor 44 as necessary, and an operator can input various
`instructions to the computer 18 from a keyboard 46.
`[0021] In readiness for cases in which the communication with
`the reference station by the communication device 40 goes bad,
`in this working example, as shown in FIG. 2, the measured GPS
`correction data is saved on a floppy disk independently at the
`reference station as well. That is to say, the GPS reception
`device 50 calculates the current location of the GPS antenna 52
`from the output of the GPS antenna 52, and outputs the GPS
`position measurement data
`to a computer 54. Accurate
`coordinates (reference position data) of the GPS antenna 52 are
`measured beforehand, and this data is input to or set in the
`computer 54. The computer 54 computes the error of the GPS
`position measurement data from the GPS reception device 50
`and the reference position data, and records it on a floppy disk
`56 as GPS correction data. Of course, measurement-time
`information is also recorded at the same time. The GPS position
`measurement data and the error (which is to say, the GPS
`correction data) are displayed on the screen of the monitor 58 as
`necessary. The operator can input various instructions to the
`computer 54 by way of a keyboard 60. The computer 54 also
`sends the GPS correction data via a communication device 62 to
`(the computer 18 of) the airborne measurement system shown in
`FIG. 1.
`[0022] The data measured by the airborne measurement system
`shown in FIG. 1 (and as necessary by the on-the-ground
`measurement system shown in FIG. 2) is analyzed, and
`three-dimensional data is calculated, by the on-the-ground
`analysis system shown in FIG. 3. That is to say, the high-quality
`VTR 70 plays back the videotape that was recorded by the
`airborne measurement system shown in FIG. 1, applying the
`image image [sic] signal to the frame buffer 74 and the
`reproduced time codes to an engineering workstation 76. The
`image data temporarily stored in the frame buffer 74 is applied to
`a monitor 78 and
`is
`image-displayed. Needless
`to say,
`sometimes the reproduced time codes are also displayed
`simultaneously on the monitor 78.
`[0023] A personal computer 80 reads the flight data and GPS
`correction data that is simultaneously collected by the airborne
`measurement system shown in FIG. 1 (in the event of a
`communication breakdown, the GPS correction data measured
`by the on-the-ground measurement system shown in FIG. 2),
`corrects the GPS position measurement data with the GPS
`correction data and corrects the tri-axial gyroscopic data with the
`magnetic bearing data, and transfers this, along with other
`measurement data and the time codes that are recorded together,
`to the workstation 76. The workstation 76 controls the VTR 70 in
`accordance with the time codes supplied from the computer 80
`and causes the VTR 70 to play back images with the same time
`code. In this way, the workstation 76 can correlate the conditions
`and imaging position when imaging with the images captured at
`that time, and reproduces three-dimensional data by way of
`computation that is described in detail below.
`[0024] FIG. 4 shows the flow in this working example, from
`measurement to three-dimensional data extraction. First, the
`equipment shown in FIG. 1 is put on board the aircraft, the target
`region is imaged and the flight information is recorded while flying
`over the target region at as constant an altitude and speed as
`possible (S1). At this time, the imaging target basically moves
`perpendicular to the scan lines of the camera 10. The images
`captured by the camera 10 are recorded on videotape by the
`VTR 14. At the same time, information on the exact position
`(latitude, longitude, height) and orientation of the camera 10 is
`
`
`Translation by Patent Translations Inc. 1-800-844-0494 mail@PatentTranslations.com
`
`
`
`
`
`JP-08-159762-A Page 6
`other so that the scan lines do not overlap, and are displayed in
`alternation every 1/60 second.
`[0035] That is to say, the video images captured by the camera
`10 are made up of scenes (fields) in which the imaging position is
`changed every 1/60 second; in this working example, as shown
`in FIG. 9, line data for the leading line, the middle line, and the
`final line of each field are extracted therefrom. The image formed
`from the data of the leading line of each field is called the forward
`view image, the image formed from the data of the middle line of
`each field is called the nadir view image, and the image formed
`from the data of the final line of each field is called the rearward
`view image. If the travel speed when imaging is constant, and the
`orientation of the camera is also constant, stereoscopic viewing is
`possible using this forward view image, this nadir view image,
`and this rearward view image. Note that lenses of short focal
`length may be used to emphasize height, which is to say, to
`increase the resolution.
`[0036] However, in video imaging with an aircraft, there are the
`variable factors of changes in the flying speed, flight path
`deviations, changes in the flying altitude, and changes in the
`three axes (pitch, roll, and yaw), and it is necessary to remove
`the vertical parallax that arises from these influences (S4).
`[0037] In the vertical parallax removal processing (S4), first,
`based on the exterior orientation elements of the images with a
`60% overlap percentage obtained in the orientation calculation
`(S2),
`the values of
`the exterior orientation elements
`corresponding to each of the line data in the continuous mosaic
`images (forward view image, nadir view image, and rearward
`view image) are interpolated. For example, as shown in FIG. 10,
`where the exterior orientation elements, which is to say, the
`position and orientation of the camera, at three imaging points P,
`Q, R, are, respectively, (Xp, Yp, Zp, ωp, ϕp, κp), (Xq, Yq, Zq, ωq,
`ϕq, κq), and (Xr, Yr, Zr, ωr, ϕr, κr), in the continuous mosaic
`image, the values of these exterior orientation elements are
`assigned to the lines that correspond to the nadir view image at
`the imaging points P, Q, R, and interpolated values are assigned
`to other lines. In this way, as shown in FIG. 11, exterior
`orientation