JPH034106A - Three-dimentional measuring method and device using video - Google Patents

Abstract

PURPOSE:To obtain variable-focus and large-scale detail information by fetching infor mation on a desired measured point from a reproduced screen which is made still, performing the arithmetic processing of the fetched information and obtaining the length, the area and the volume of the desired measured point. CONSTITUTION:A photographic tape is dubbed and reproduced by two video decks 2 and 3 with a time refference which gives a desired base length, then alternately displayed on the screen of a monitor 1 by a switcher 6. In order to obtain the measured value of the distance between spots and the area, etc., concerning a desired spot from the video screen which is obtained as a stereoscopic pair, a ground distance D at the time of photographing is obtained. The distance D is obtained on a map based on the flight course and the flight altitude (h) of a helicopter. Then, a focal distance (f) is obtained by reading a value imprinted on a video and photographing scale is calculated from the distances D and (f). By displaying the video of a collapsed place on the screen of the monitor 1, obtaining the area on the screen by the use of a cursor on the screen and correcting the value by using the photographing sale, an actual collapsed area is calculated.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention is useful for monitoring and determining collapsed areas in mountainous areas, trees on power transmission lines in mountainous areas, snowfall conditions in mountainous areas, ocean current conditions during stormy weather, etc. The present invention relates to a suitable three-dimensional measurement method and apparatus.

Investigations of collapsed areas in mountainous areas that conventional engineers cannot easily access, trees on power transmission lines in mountainous areas, snowfall in mountainous areas, and ocean current conditions during stormy weather when ships cannot be launched are carried out through erosion control management, planning, and cutting down of obstructing trees. Provides important information for comprehensive disaster prevention and construction planning, including planning, power transmission line construction plans, water resource management plans, and coastal revetment construction plans. For this purpose, necessary information has conventionally been obtained through photogrammetry surveys using aerial photographs and ground surveys by humans.

However, with conventional photogrammetry using aerial photographs, only extremely small-scale photographs can be obtained due to the flight altitude and camera focal length. In addition, in steep mountainous areas, valleys are often in shadow and cannot be seen in the photograph itself, and the survey is only a general survey. In addition, in mountainous areas, severe weather conditions often restrict the flight of aircraft, and there are many areas that are impossible to explore even from the ground. In the case of phenomena such as ocean currents where it is necessary to capture temporal changes, continuous imaging over a long period of time is necessary.

Meanwhile, society's demands have changed from general surveys using small-scale aerial photographs to detailed surveys using large-scale detailed information, and these demands have also changed from information sources that capture a momentary moment in time, such as photographs, to continuous information sources. The research methods used have changed. In order to meet these recent demands, the present inventors previously developed a solid image monitoring method using video images as disclosed in Japanese Patent Application Laid-Open No. 59-85909, and It has been applied to research. This real image monitoring method is
Continuous images along a desired route, such as a power line route, are created at a nearly constant speed by vertical photography from a nearly constant height using a video camera, etc., and a copy image is created using the continuous images as the original image, and both images are are displayed simultaneously and consecutively with a time difference that causes parallax, so that the display from one of the two images is visible only to one eye of the observer, and the display from the other of both images is visible to the observer's other eye. It should only get into the eyes.

Problems to be Solved by the Invention According to the solid image monitoring method using video images as described above, the basic original images are recorded at a ground speed of 1100 K/2.
If the observation is carried out using a helicopter around the same time, it will be possible to cover about 200Kfll of power transmission line routes in one operation time (about 2 hours), making it possible to easily conduct observations over a wide area in an extremely short period of time. Moreover, since the observation is carried out by monitoring continuous solid images, it is possible to obtain information equivalent to or better than that obtained from photographs on photographic paper or human field surveys. However, with this solid image monitoring method, it is only possible to judge the situation by visually viewing continuous solid images, and it is not possible to obtain the distance between desired points or the area or volume of the desired place using numerical measurement values. Therefore, it was difficult to grasp the situation at a desired point as numerical measurement value information.

An object of the present invention is to provide a three-dimensional measurement method and apparatus using video images that can solve these conventional problems.

Means for Solving the Problems According to the three-dimensional measurement method using video images of the present invention,
A video camera captures continuous images along the desired route to be measured, and the captured video images are played back with a time difference that creates a parallax to form a stereo pair, and a playback screen including the desired measurement points is created. The device is made to stand still, captures information regarding the desired measurement point from the frozen playback screen, and calculates the length and area of the desired measurement point by processing the captured information.
Obtain measurement values such as volume.

Furthermore, the three-dimensional measuring device using video images of the present invention includes:
a video playback device equipped with a display for playing back video footage shot continuously along a desired route to be measured; and a static playback screen containing the desired measurement points with a time difference that creates a stereo pair of parallaxes; a control means for controlling the reproducing device to alternately display on the display, an information importing means for importing information regarding the desired measurement point from the static reproduction screen, and arithmetic processing of the retrieved information. The length and area of the desired measurement point,
and arithmetic processing means for obtaining measured values such as volume.

Embodiments Next, the present invention will be described in more detail with reference to embodiments of the present invention based on the accompanying drawings.

FIG. 1 is a block diagram showing an embodiment of a three-dimensional measuring device using video images according to the present invention. As shown in FIG. 1, this device includes a monitor 1 for displaying video images shot with a video camera as described later.
, two video decks 2 and 3 for playing video tapes, a keyboard 4 for manually inputting shooting parameters such as ground distance and focal length, and a keyboard 4 for manually inputting shooting specifications such as ground distance and focal length; A mouse 5 for reading out various information as described later from a desired point, and a switcher 6 for alternately switching between the image signal reproduced by the video deck 2 and the image signal reproduced by the video deck 3 and inputting the same to the monitor 6. , and a printer 7 for printing out measurement results. Monitor 1 is CPU, R
It has a built-in arithmetic processing unit including OM and RAM, and various interfaces.

An embodiment of the measuring method of the present invention will be described below with particular reference to the schematic diagram in FIG. 2 and the flowchart in FIG. do.

First, in step 10 in the flowchart of FIG. 3, as schematically shown in FIG. 2, a video camera mounted on a helicopter photographs the collapsed land in a horizontal direction. The flight route is set in advance so that the optical axis is perpendicular to the collapse surface when photographing each collapse site. Also, shoot so that the optical axis is perpendicular to the aircraft. Furthermore, the date and time, altitude above the ground, and focal length are reflected in real time in the footage taken as a tirop.

Next, in order to carry out desired measurements from the video images taken by the video camera in this manner, in step 11 in FIG. 3, a stereo pair of images including the point to be measured is created. In this stereo pair, an arbitrary baseline length can be obtained by giving a time difference to the images shot by one video camera. In order to obtain a left and right stereo pair, two video tapes are actually created by dubbing the photographic tape, and the desired baseline length is created using the two video decks 2 and 3 in the apparatus shown in FIG. The images are reproduced with a given time difference and are alternately displayed on the screen of the monitor 1 by the switcher 6.

In order to obtain desired measurement values such as distance, area, volume, etc. between points regarding a desired point from the stereo pair video screen obtained in this way, first, the distance to the ground at the time of shooting must be determined. However, how to determine the distance to the ground will be explained with particular reference to FIG. Fourth
As shown schematically in the figure, the distance to the ground is
8 flight course and flight altitude can be determined on the topographic map. The flight course is recorded vertically downward using a video camera fixed vertically to the aircraft, and the flight path is shown on a topographic map from the images. The flight altitude at the time of shooting is determined by reading the value reflected in the video image.

The distance to the collapsed ground is determined by the following formula according to the six values shown in Fig. 4.

Distance to the ground D = (h+H2-Hl) 2+L'Next, the focal length f is obtained by reading the value reflected in the video image.For example, when shooting during measurement, it is fixed at 12.49u. ing.

From the distance to the ground thus obtained and the focal length f, first, in steps 12, 13, and 14 of FIG. 3, the photographing scale is calculated. The details of the flow for determining the imaging scale are shown in FIG. 1O. Prior to calculating the photographing scale on the screen from the given ground distance and focal length f, the camera is tested in advance to determine the correlation between the ground distance and the photographing scale. Its distance from the ground D
The relationship between and the photographing scale is illustrated in FIG. Since the photographing scale is measured on the screen, the basic unit is one dot. From the relationship shown in Figure 5, focal length f =
The correlation line between the two at 12.49 mm is expressed by the following equation.

Photography scale (m/dot) = 1.04 x l O-'xD + 4.68 x 10
-3 However, this relational expression is valid only for images captured and displayed by the video camera and monitor used in this embodiment. To calculate the photographing scale on the screen, as shown in the flowchart in FIG. In step 102, calculation is performed using the above-mentioned correlation formula, and in step 103, the imaging scale is calculated in the order in which it is calculated.

The video camera uses a gyroscopic mechanism called Wescam System (product name) to keep the optical axis parallel when shooting. However, since there is a possibility that distortion may occur in the left and right stereo pair due to the influence of the flight direction of the aircraft, etc., it is necessary to perform correction using affine transformation. Therefore, as shown in step 15 in FIG. 3, in the paired images, the cursor is moved using the mouse 5 in the device shown in FIG. Based on the coordinates of these three points, and using one image as a reference, in step 16 of FIG.
Calculate Sb,c. The three reference points set here are
Points around the collapsed area, and at the same time, the surface determined by these three points is used as a virtual surface that serves as a reference for determining the collapse depth (see Figure 4). These measurements are performed using a mouse cursor superimposed on the still screen displayed on the monitor 1 in the apparatus shown in FIG. 1 by operating the mouse 5.

The calculation of the affine coefficients is performed by solving simultaneous equations based on the given coordinates of three reference points on the left and right sides, as will be explained in detail later.

Next, in step 17 of Fig. 3, it is necessary to input the photo baseline length P, which is the distance between the arbitrarily determined left and right principal points of the stereo pair. Calculated as a number. Also, as shown in step 18 in Figure 3, it is necessary to input the desired points using the cursor, but this is done by selecting the deepest points within the collapse area and placing them on the left and right images of the stereo pair. Manually mark the corresponding points using the cursor.

Furthermore, as shown in step 19 of FIG. 3, it is necessary to calculate the parallax difference, but this is done by calculating the parallax difference on the images that are the stereo pair, as shown schematically in FIG. 6. The coordinates of the sought point are multiplied by the affine coefficient obtained in step 16 and converted onto the reference image. At this time, the deviation between the coordinates of the sought point on the reference plane and the coordinates of the sought point after conversion becomes the sought parallax difference ΔP.

In step 20 of Fig. 3, the collapse depth d is calculated, which is calculated using the collapse base point obtained in step 15, the photographic base line length obtained in step 17, and the collapse depth d obtained in step 19. It is obtained from the following relational expression using each numerical value of the parallax difference.

Next, in step 21 of Fig. 3, the collapse area A is measured, but this is done by displaying an image of the collapse area on the screen of the monitor l using the method described later. This is done by determining the area on the screen using a cursor and correcting it based on the photographic scale to calculate the actual area of collapse.

Also, in step 22 of Fig. 3, the amount of collapsed soil V is calculated, which is calculated from the collapse depth d obtained in step 20 and the collapse area A obtained in step 21. .

The photographing specifications, measurement results, etc. thus obtained can be outputted by the printer 7 in the apparatus shown in FIG.

Next, before explaining a more specific example of measurement using the method of the present invention, the accuracy of measurement using the method of the present invention will be explained.

Measurement accuracy depends on the distance to the ground at the time of photography, the focal length of the lens, the number of dots on the monitor, etc. In this embodiment, the x and y coordinates of an arbitrary point are measured by placing a mouse cursor on the screen on the monitor, so the degree of resolution depends mainly on the number of dots on the monitor. In this embodiment, when superimposing, 51
It has 2x512 dots. Therefore, suppose 8.8
m + s
49a+a+ field 8, 8mm x (200m/12.
49mm) X (11512) =0.2811
1, one dot in the X direction corresponds to 0.28 m.

Similarly, one dot in the y direction is 6.6mm x (200m/12.49m+n)
X (11512) =0.21m. 1 in 2 directions
The dot is expressed by the formula, and when the ground road fiD=constant, d is
It depends on the screen baseline length P. Figure 7 shows ground distance and baseline length P.
One dot in two directions is shown when various changes are made. This seventh
Looking at the graph in the figure, the longer the baseline length (the shorter the overlap), the shorter the distance to the ground, the smaller the length per dot, and the better the resolution. However, increasing the baseline length and shortening the ground distance have contradictory properties during actual photographing. That is, in order to obtain a long baseline length, the photograph must be taken at a position far from the collapsed site. In this case, the distance from the ground becomes longer and the resolution deteriorates. Furthermore, as the object becomes smaller, measurement errors during measurement become larger. On the other hand, when the distance from the ground is reduced, the subject appears larger, but the baseline length becomes shorter. There is also a high risk that the collapsed area being photographed will cover the screen. For example, if there are large and small landslides in succession, it is better to reliably reflect the landslides, even if the distance from the ground is longer, so that you can grasp the surrounding situation necessary for the reference point.
It is safe at the same time. Therefore, the flight route was set so that 80% of the baseline length could be obtained for each collapsed site before photographing, and the aiming measurement accuracy was set at 0.5 m.

Next, in order to make it easier to understand the measurement method of the present invention in more detail, a simple example of measuring the volume of a cylinder shown on a video image will be described.

To measure the volume of a cylinder that appears on a video image, it is first necessary to find the area of the top surface of the cylinder. For information on how to find the area of this circular surface, refer to Figures 8 and 12. explain.

First, as shown by the dotted line in Figure 8 (^), an image of a circular surface to be measured is reproduced and displayed in a stationary state on the screen of the monitor l. Next, in step 108 in the flowchart for calculating the area shown in FIG. 12, the area object is captured on the screen using the cursor and manually entered into the memory within the monitor 1.

As shown in FIG. 8(B), the information regarding the area object is captured by the cursor at the points C,,C,...
...This is performed by tracing the cursor along the target arc by operating the mouse 5 as shown in C7, and reading and inputting the X1Y coordinates of each point on the screen. After reading the X and Y coordinates of each point on the target arc in this way, first calculate the area a of the target circle on the screen as shown in Figure 8 (C). In step 109, calculation is performed from the imported X and Y coordinate values. From the area a on the screen calculated in this way, the actual area A is calculated from step 1 in Figure 12.
It can be corrected in step 10 based on the tilt of the screen and the photographing scale, and then determined in step ill.

Next, a method for determining the volume of a cylinder on a video image will be described with reference to FIGS. 9 and 13.

First, in step 1'12 in the flowchart for calculating the volume in FIG. 13, as shown in FIG. 9(A), the monitor l is
A switcher 6 alternately switches and displays static playback video images from the video decks 2 and 3 as the right screen and the left screen of the stereo pair on the screen of the stereo pair. Then, in step 113 in FIG. 13, the reference point is manually entered into the memory in the monitor 1 using the mouse 5 and the cursor on the right screen. As shown in the right photo of Figure 9 (^), this means that three reference points (x+, yl), (X2, y2), (
This is done by reading each position (coordinate value) of X3, V3) with the mouse cursor. These three reference points are assumed to be on the same plane.

Then, in step 114 of FIG. 13, move the reference point on the left screen to the monitor 1 with the cursor
input into memory within. As shown in the left photo of Fig. 9 (^), this means that three corresponding reference points (X,, Y,), (X2, Y2), ・(
This is done by reading each position (coordinate value) of X3, Y3) with one mouse cursor.

Next, from the xSy coordinates obtained from the three left and right reference points input as described above, the geometric correction coefficient (affine coefficient) from the right screen to the left screen (or from the left screen to the right screen) is calculated.
is calculated using the following simultaneous equations.

X1=AXi+BYi+C Yi=DXi+EYi+F (A-F: unknown quantity) This coefficient corrects distortion in the left and right planes.

Next, input the measurement points (points with different relative heights) for which you want to find the relative height. For this purpose, in step 115 of FIG. 13, the user inputs the desired intention using the cursor on the right screen, and in step 116, inputs the desired intention using the cursor on the left screen. That is,
Use the mouse cursor to read the coordinates of point P' on cylinder 9' in the right photo of FIG. 9(B), and use the mouse cursor to read the coordinates of point P' on cylinder 9 in the left photo of FIG. 9(B). read.

Next, in step 119 of FIG. 13, a parallax difference is calculated. In steps 115 and 116, the measurement points input on the left and right sides are affine-transformed from right to left (or from left to right) using the affine coefficients obtained as described above. Here, since the converted measurement point has a different relative height from the plane defined by the reference point, a shift (parallax difference ΔP) occurs in the converted coordinates. Furthermore, step 11 in FIG.
In step 7, calculate the baseline length Po on the screen, and proceed to step 1.
In step 18, the distance to the ground H is calculated, and then, in step 120, the photo base line length P obtained in the previous step is calculated.
The relative height is calculated from the O1 ground distance aH and the parallax difference ΔP using the following formula.

Specific height = (ΔP/ΔP+Po) xH Finally, the area A of the circular surface of the cylinder obtained as described above
is output in step 121 and the specific height determined in step 120 allows the volume V of the cylinder to be determined in step 122 using the following equation.

Volume V = Area AX ratio height Figure 11 shows 1. This is a flowchart for calculating the length between two points from a still video image displayed on the screen of a monitor l, and as shown in FIG. 11, each step 104.10
5. Through steps 106 and 107, the length between two points can be measured using a method similar to the method used to determine the volume of the cylinder described above.

Effects of the Invention According to the three-dimensional measurement method and device using video images according to the present invention as described above, all of the following problems encountered when using vertical photographs of fixed-wing aircraft and field surveys as in the past can be solved. can do.

(1) Low flight potential of visual flight fixed-wing aircraft in mountainous areas.

(2) Existence of areas in mountainous areas where it is impossible to take vertical photographs of fixed-wing aircraft.

(3) Small-scale vertical photography of fixed-wing aircraft over mountainous areas.

(4) Information that is momentary in time.

(5) Existence of areas that cannot be explored by humans.

(6) High cost and long construction period for long-distance and wide-ranging survey targets.

In addition to these effects, the present invention also provides the following effects.

(1) Detailed information on variable focus and large scale can be obtained.

(2) Continuous information can be obtained.

(3) Photographs are not required.

(4) It is easy to collect information for investigation.

(5) Measurements can be easily performed on a video screen.

(6) By storing the original video footage and the duplicate video footage in a data file, they can be played back as many times as necessary for investigation, interpretation, and measurement.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of a three-dimensional measuring device using video images of the present invention, FIG. 2 is a schematic diagram for explaining the video image capturing method of the present invention, and FIG.
4 is a schematic diagram for explaining the measurement of distance to the ground in measurement according to the present invention. FIG. 5 is a diagram showing the distance to the ground in measurement according to the present invention. FIG. 6 is a diagram for explaining the affine transformation in measurement according to the present invention, and FIG. 7 is a diagram showing the relationship between height above ground and height when the baseline length is changed in measurement according to the present invention. A diagram showing the relationship of resolution, Figure 8, is
Figure 9 is a diagram for explaining the procedure for determining the area of a circular surface by measurement according to the present invention. Figure 1O is a diagram for explaining the procedure for determining the volume of a cylinder by measurement according to the present invention. FIG. 11 is a diagram showing the flow for determining the scale in measurement according to the invention. FIG. 12 is a diagram showing the flow for determining the length between two points in measurement according to the invention. The figure shown in FIG. 13 is a diagram showing a flow for determining a volume in measurement according to the present invention. 1...Monitor, 2.3...Video deck, 4...Keyboard, 5...Mountain/mouse, 6...Switcher 7... ...Printer, Death...Helicopter-9,9'...River/Cylinder. Figure 3 Figure 2 Figure 4 Figure 5 100 200 Distance to ground 0 (m) 00 Figure 8 B Figure 6 Figure 7 Distance to ground (m) Figure 9 (B) Standard length Figure 10 Figure 11 Figure 12 Flow for calculating area

Claims (2)

[Claims] (1) Shoot continuous images along the desired route to be measured with a video camera, play back the shot video images with a time difference that creates a parallax to form a stereo pair, and include the desired measurement point. The playback screen is frozen, information regarding the desired measurement point is captured from the frozen playback screen, and the captured information is processed to determine the length of the desired measurement point.
A three-dimensional measurement method using video images, which is characterized by obtaining measured values such as area and volume. (2) A video playback device equipped with a display for playing back video footage shot continuously along the desired route to be measured, and including the desired measurement points with a time difference that creates a stereo pair parallax. a control means for controlling the playback device so as to alternately display still playback screens on the display; an information fetching means for fetching information regarding the desired measurement point from the still playback screen; and a calculation for the fetched information. the length of the desired measurement point after processing;
1. A three-dimensional measuring device using video images, characterized by comprising arithmetic processing means for obtaining measurement values such as area and volume.