Camera orientation method using agricultural sheets

Abstract

To reduce labor required for three-dimensional photo measurement of agricultural products.SOLUTION: An agricultural mulch film 100 covering a ground surface planted with plants comprises a sheet-shaped member and a plurality of identification markers 101-116 for image recognition used for three-dimensional photo measurement arranged on the surface of the sheet-like member. The agricultural mulch film 100 is arranged in a state of covering a ridge 200 in a field. Some of the identification markers 101-116 are positioned on the upper part of the ridge 200 and others are positioned on the slope of the ridge 200.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an agricultural sheet for covering soil on agricultural land and a technique using the agricultural sheet. [Background technology]

[0002] Agricultural sheets that cover the ground surface in agricultural fields are known (see, for example, Patent Document 1). The agricultural sheets have holes drilled here and there, through which plants are exposed for cultivation. The agricultural sheets are used to suppress the growth of weeds and to retain heat. The agricultural sheets are also called agricultural mulch films, mulch sheets, weed-control sheets, ground sheets, etc.

[0003] Incidentally, a technique for using three-dimensional photographic measurement in growing agricultural crops is also known (see, for example, Patent Document 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Publication No. 58-068849 [Patent Document 2] Japanese Patent Application Laid-Open No. 2004-191243 Summary of the Invention [Problem to be solved by the invention]

[0005] When performing three-dimensional photogrammetry, a target is required for orientation. The exterior orientation parameters (position and orientation) of the camera are determined. For example, aerial photogrammetry involves placing a number of orientation targets on the Earth's surface.

[0006] The placement of orientation targets is also necessary when performing three-dimensional photometry of plants. When placing orientation targets, it is desirable to place multiple orientation targets as three-dimensionally (stereoscopically) as possible. In addition, in order to give scale to the relative three-dimensional model obtained by relative orientation, the distance between the two or more orientation targets used must be known.

[0007] The installation of orientation targets is a complicated task, and simplification of the process is desired. In this context, the present invention aims to provide a technology that can reduce the amount of work required when performing three-dimensional photo measurement of agricultural products. [Means for solving the problem]

[0008] The present invention uses an agricultural sheet to cover the ground where plants are planted. Camera Orientation A method, comprising: the agricultural sheet comprising a sheet-like member; The aforementioned and a plurality of image recognition markers used for camera orientation and arranged on the surface of the sheet-like member, wherein the ground has upwardly raised ridges formed thereon, and the ridges are covered with the agricultural sheet, and the agricultural sheet in a state of covering the ridges has a high portion which is a portion of the upper surface of the ridges, a low portion which is located lower than the high portion and is covered with soil, and a slope portion which is a slope between the high portion and the low portion, and one of the plurality of image recognition markers is arranged on the high portion and another of the plurality of image recognition markers is arranged on the slope portion, and placed in the higher ranks Image recognition markers an image recognition marker disposed on the slope portion; The camera is oriented using an agricultural sheet. Camera Orientation It is a method.

[0009] In one embodiment of the present invention, the plurality of image recognition markers are individually identifiable and their central positions can be detected from an image. In another embodiment of the present invention, a growing position is determined for each plant to be grown on the sheet-like member, and a correspondence relationship is specified between at least one of the plurality of image recognition markers and the growing position.

[0010] In one embodiment of the present invention, the plurality of image recognition markers include at least two image recognition markers spaced a known distance apart. In another embodiment of the present invention, the agricultural sheet covers ridges, and at least some of the plurality of image recognition markers are arranged at positions corresponding to portions of the ridges that have differences in elevation.

[0011] In one embodiment of the present invention, the plurality of image recognition markers are arranged separately in a portion corresponding to the upper part of the ridge and a portion not corresponding to the upper part. In one embodiment of the present invention, when the ridges are covered, soil is placed around the edges of the agricultural sheet, and the plurality of image recognition markers are arranged to avoid the portion covered with soil. In one embodiment of the present invention, the sheet-like member has a random dot pattern. [Effects of the Invention]

[0013] According to the present invention, the amount of work required to perform three-dimensional photo measurement of agricultural products can be reduced. [Brief explanation of the drawings]

[0014] [Figure 1] 1A is a top view and FIG. 1B is a cross-sectional view of an agricultural mulch film using the present invention installed in a field. [Figure 2] FIG. 1 is a diagram illustrating an example of an imaging system. [Figure 3] 10 is a flowchart illustrating an example of a processing procedure. [Figure 4] FIG. 1 is a diagram illustrating the principle of orientation. [Figure 5] FIG. 1 is a diagram illustrating the principle of forward intersection. [Figure 6] FIG. 10 is a top view of another agricultural mulch film using the present invention installed in a field. [Figure 7] 1A is a top view and FIG. 1B is a cross-sectional view of another agricultural mulch film using the present invention installed in a field. DETAILED DESCRIPTION OF THE INVENTION

[0015] 1. First embodiment (Overall Overview) Fig. 1 shows an agricultural mulch film 100. Fig. 1(A) shows one of the ridges in a farm field covered with the agricultural mulch film 100, as viewed vertically from above. Fig. 1(B) is a cross-sectional view of the ridge covered with the agricultural mulch film, as viewed from the direction of extension of the ridge.

[0016] The agricultural mulch film 100 is used to cover the ground of agricultural land. The agricultural mulch film 100 is a thin sheet-like member made of resin with a thickness of about 0.01 mm to 0.5 mm.

[0017] Here, the agricultural mulch film 100 used is made of black polyvinyl chloride with a thickness of 0.02 mm. Materials that can be used for the agricultural mulch film 100 include polyvinyl chloride, polyolefin, fluororesin, polyethylene, etc. Woven fabric and nonwoven fabric can also be used. The light (sunlight) transmittance, moisture permeability, and water permeability can be selected as appropriate. The physical properties required of the material are the same as those of commercially available agricultural sheets.

[0018] The agricultural mulch film 100 has a strip-like longitudinal shape with a width of 1 m to 2 m and a length of several tens of meters or more. The agricultural mulch film 100 is rolled up into a cylindrical shape, and when in use, it is unrolled and cut to an appropriate length. This is the same as commercially available agricultural sheets.

[0019] FIG. 1 shows a state in which a ridge 200 formed in farmland is covered with an agricultural mulch film 100. In FIG. 1, the Y direction is the longitudinal direction (extension direction) of the agricultural mulch film 100, and the X direction is the width direction. The Z direction is the vertical direction. FIG. 1 shows a portion of the longitudinal direction of the agricultural mulch film 100. In an actual farm field, the state shown in FIG. 1(A) exists along the extension direction of the ridge 200 for the length of the ridge 200.

[0020] Identification markers 101 to 116 are printed on the surface that will be on the top when the agricultural mulch film 100 is installed. It is also possible to have the identification markers 101 to 116 attached to the surface of the agricultural mulch film 100. It is also possible to place identification markers on both sides of the agricultural mulch film 100 so that it can be used regardless of the front or back.

[0021] The identification markers 101 to 116 are targets for image recognition that can be individually identified by image recognition. The code data displayed on the identification markers 101 to 116 can be marks of various shapes, a combination of multiple figures, a form using character data, a two-dimensional barcode, etc.

[0022] In this example, the center positions of the identification markers 101 to 116 can be identified from the image. These center positions are the positions of the identification markers 101 to 116.

[0023] Specifically, at the center of each of the identification markers 101 to 116, there is a circle indicating that it is the center, and in addition, a figure with curvature surrounding the center is arranged around the circle. The position of the center can be calculated from this figure with curvature. For example, the position of the curvature center of the figure is set to coincide with the center of the identification marker. Furthermore, the combination of these figures constitutes a code for individually identifying the identification marker.

[0024] The spacing between the identification markers 101 to 116 in the longitudinal direction (Y direction) of the agricultural mulch film 100 is the same distance L1. L1 is a known value that is determined in advance. It is also possible to make the distance between the identification markers in the Y direction not all the same. In this case, the combination of adjacent identification markers and the distance between them are stored as known data. The distance between adjacent identification markers is measured based on the center position of each identification marker.

[0025] The spacing between the identification markers 101 to 116 in the width direction (X direction) of the agricultural mulch film 100 is the same distance L2. L2 is a known value that is determined in advance. It is also possible for the distances between the identification markers in the X direction not to be all the same. That is, L2 may be the same value or different values. L1 and L2 may be the same value or different values.

[0026] Holes (openings) 121-123 are provided at the center line in the longitudinal direction (the direction of the Y axis in FIG. 1) of the agricultural mulch film 100, for allowing the buds and stems of the plants being cultivated (for example, the plant designated by reference numeral 300) to exit. The holes 121-123 are provided at equal intervals, and the distance L3 between them is approximately 10 cm to 50 cm. This distance varies depending on the plants being cultivated. There are no particular limitations on the type of plants being cultivated.

[0027] These holes 121 to 123 are the growing positions for each individual plant. In this example, the overall design is such that four identification markers surround each of the holes 121 to 123. For example, hole 121 is surrounded on all four sides by identification markers 102, 103, 106, and 107.

[0028] The positional relationship between the identification markers 101-116 and the holes 121-123 is known. Therefore, for example, once the positions of the identification markers 102, 103, 106, and 107 are identified, the position of the hole 121, i.e., the position of the plant cultivated there, can be identified. Furthermore, cultivated crops can be managed on an individual basis. For example, management can be performed such that individual crop A is surrounded by identification markers 102, 103, 106, and 107, and individual crop B is surrounded by identification markers 106, 107, 110, and 111.

[0029] When the agricultural mulch film 100 covers the ridge 200, the first group of identification markers, 102, 103, 106, 107, 110, 111, 114, and 115, are located on the approximately flat upper surface of the ridge 200. The second group of identification markers, 101, 104, 105, 108, 109, 112, 113, and 116, are located on the slopes of the ridge 200.

[0030] The first group of identification markers and the second group of identification markers are positioned relative to each other with a difference in elevation. As a result, the identification markers 101 to 116 are arranged in the horizontal direction with the distribution shown in Figure 1(A), and in the vertical direction with the difference in elevation shown in Figure 1(B). By arranging the identification markers three-dimensionally, the accuracy of three-dimensional photometry can be improved.

[0031] In addition to the identification markers printed on the agricultural mulch film 100, an identification marker 211 is also placed on top of a stake 210 that is erected at the top of the ridge 200. As the cultivated plants 300 grow, there is a possibility that the identification marker will be hidden by leaves or the like. Placing the identification marker at a higher position on the stake 210 alleviates this problem.

[0032] The height of the identification markers 211 is set higher than the height of the plants 300 to be grown. The number of identification markers placed at this higher position is set to about one for every four to eight identification markers in the two central rows.

[0033] To prevent the agricultural mulch film 100 from being turned over or blown away by the wind, soil is covered on the left and right longitudinal edges of the agricultural mulch film 100. Reference numerals 131 and 132 indicate the portions that are covered with soil from above.

[0034] If the identification markers 101 to 116 are covered with soil, their functionality will be impaired, so the identification markers 101 to 116 are arranged to avoid the portions of the agricultural mulch film 100 that will be covered with soil.

[0035] (Method of shooting) One method of taking photographs is to use a moving object. In this method, a camera is mounted on the moving object and the object is moved along the length of the furrows while repeatedly taking photographs with the camera. The moving object can be one that moves on wheels or caterpillars, one that moves along rails or wires, one that moves on wires, or a UAV (unmanned aerial vehicle).

[0036] For example, a moving object is moved along the ridges at a speed of 10 cm / sec, and photographs are taken repeatedly with a camera every second. In this case, many pairs of stereo photographs are obtained, each pair taken from adjacent viewpoints (camera positions) that are 10 cm apart. Note that the overlapping areas of the obtained stereo photographs are the subject of stereo photographic measurement, so the photographing conditions should be adjusted with this in mind.

[0037] It is also possible to mount two cameras as stereo cameras on a moving object in advance, and use the two cameras to capture stereo images while moving the moving object.

[0038] The number of cameras mounted on the moving object may be two or more. For example, it is possible to prepare a camera that takes pictures from a high angle position, a camera that takes pictures from a medium angle position, and a camera that takes pictures from a horizontal direction.

[0039] In either case, a stereo image is obtained from the multiple images obtained, and the shooting conditions are set so that no images are missed. Also, images are taken from as many different directions as possible.

[0040] It is also possible for a person to hold a camera and take pictures. In this case, two or more pictures are taken of the same subject by changing the viewpoint (camera position). The conditions for taking pictures are the same as when using a moving body. It is also possible to use multiple fixed cameras (including rotatable cameras).

[0041] An example of a photography system is shown in Figure 2. The viewpoint in Figure 2 is the same as that in Figure 1(B). Figure 2 shows carts 501 and 503 moving on both sides of ridge 200 in the extension direction of ridge 200. Cart 501 has a support 502, and cart 503 has a support 504. The tops of support 502 and 504 are connected by beam 505. Cameras 506, 507, 508, and 509 are held on beam 505.

[0042] When viewed from the viewpoint of Figure 2, with the position of plant 300 as the base point, camera 506 is positioned at a 45° angle upward to the left, camera 507 at a 60° angle upward to the left, camera 508 at a 60° angle upward to the right, and camera 509 at a 45° angle upward to the right.

[0043] When taking the photographs, the carts 501 and 503 are moved along the extension direction of the ridge 200 (the direction of the Y axis in Figure 1), and each camera takes photographs repeatedly at regular time intervals. It is also possible to take a video and use the frame images that make up the video.

[0044] For example, cameras 506 and 507, cameras 507 and 508, and cameras 508 and 509 are stereo cameras. Also, camera 506 that took the nth image and camera 506 that took the n+1th image are stereo cameras. In this case, the image capturing position of camera 506 is shifted in the extension direction of ridge 200 between the nth and n+1th images. Therefore, a stereo image can be obtained with one camera. This is the same for the other cameras.

[0045] (Example of processing procedure) 3 is a flowchart showing an example of a processing procedure. First, an image of the plant 300 is taken (step S101). In this example, the image is taken using the system shown in FIG.

[0046] After capturing an image, the image data of the captured image is sent to an image data processing device and read. In this example, a PC (personal computer) is used as the image data processing device. Application software that performs the processes of steps S102 to S108, which will be described later, is installed on the PC, and the processes of steps S102 to S108 are performed by the PC.

[0047] The image data processing device can be configured with dedicated hardware. The processing performed by the image data processing device can also be performed by a data processing server located at a remote location. In this case, the image data is sent to the data processing server via the Internet, and the processing of steps S102 to S108 is performed there.

[0048] When the image data processing device receives the image data, it detects identification markers from the image (step S102). The identification markers include image-readable code data, and each identification marker is identified from this code data. At this time, the position of each identification marker in the captured image is also detected.

[0049] Next, the system detects the identification information of the identification markers detected in the image and calculates the position information of the markers in the image. The camera position (position of the shooting viewpoint) is estimated (step S103).

[0050] Fig. 4 is a diagram showing the principle of estimating (calculating) the camera position (principle of orientation). Fig. 4 shows a case where identification marker 1, identification marker 2, and identification marker 3 are photographed using camera 1 and camera 2.

[0051] Here, camera 1 and camera 2 are not limited to the case where two separate cameras are used, but also include the case where one camera is used to take a picture at a first camera position (taking a picture with camera 1) ⇒ then moved to take a picture at a second camera position (taking a picture with camera 2). Note that the interior orientation parameters of camera 1 and camera 2 are assumed to be known.

[0052] In the case of Figure 4, identification markers 1, 2, and 3 appear in the image captured by camera 1, and identification markers 1, 2, and 3 appear in the image captured by camera 2. In other words, cameras 1 and 2, which are in different positions, capture images of an overlapping area, and identification markers 1, 2, and 3 are placed in that overlapping area.

[0053] Here, the position of identification marker 1 is P1, the position of identification marker 2 is P2, and the position of identification marker 3 is P3. Furthermore, the positions of point P1 on the shooting screen of camera 1 are p1, p2, and p3, and the positions of point P2 on the shooting screen of camera 2 are q1, q2, and q3.

[0054] Hereinafter, the position of camera 1 will be referred to as the position of the projection origin (the optical origin in the optical system of the camera) of camera 1. This is also true for camera 2.

[0055] Here, the direction line connecting camera 1 and identification marker 1 is A1, the direction line connecting camera 1 and identification marker 2 is A2, the direction line connecting camera 1 and identification marker 3 is A3, the direction line connecting camera 2 and identification marker 2 is B1, the direction line connecting camera 2 and identification marker 2 is B2, and the direction line connecting camera 2 and identification marker 3 is B3.

[0056] Direction line A1 is obtained as a straight line connecting the projection origin of camera 1 and position p1 on the shooting screen of camera 1. Direction line A2 is obtained as a straight line connecting the projection origin of camera 1 and position p2 on the shooting screen of camera 1. Direction line A3 is obtained as a straight line connecting the projection origin of camera 1 and position p3 on the shooting screen of camera 1. Direction lines B1, B2, and B3 are obtained in a similar manner.

[0057] Here, a triangular pyramid is obtained that is composed of the direction lines A1, A2, and A3, the line connecting P1 and P3, the line connecting P2 and P3, and the line connecting P1 and P3. That is, a three-dimensional figure is obtained whose vertices are camera 1 and points P1 to P3. At this stage, the size of this three-dimensional figure is unknown. In this way, the relative positional relationship between camera 1 and points P1 to P3 becomes clear. In other words, a relative three-dimensional model that clarifies the relative positional relationship between camera 1 and points P1 to P3 is obtained.

[0058] Similarly, the relative positional relationship between camera 2 and points P1 to P3 becomes clear. Furthermore, the line passing through the projection origin of camera 1 and the center of the photographed screen becomes the optical axis of camera 1. Similarly, the line passing through the projection origin of camera 2 and the center of the photographed screen becomes the optical axis of camera 2. Therefore, the orientations (postures) of cameras 1 and 2 relative to identification markers 1 to 3 are also determined. This is the principle of relative orientation.

[0059] Here, for example, if the separation distance between P1 and P2 is given, a scale is given to the above relative positional relationship, and the mutual positional relationship becomes clear. This is the principle of absolute orientation. As a result of this absolute orientation, the positions and orientations (postures) of camera 1 and camera 2 relative to points P1, P2, and P3 are determined.

[0060] According to the above principle, the exterior orientation parameters (position and orientation) of the camera at each camera position that photographed the farm field in Fig. 1 are calculated. This process is performed in step S103.

[0061] 4, if the positions of points P1 to P3 in the absolute coordinate system are unknown, the exterior orientation parameters of camera 1 and camera 2 that are determined will be those based on points P1 to P3. Even in this case, an absolute three-dimensional model (a three-dimensional model whose dimensions can be evaluated) of the plant to be observed can be constructed on a local coordinate system based on points P1 to P3, and changes in its shape and size can be evaluated numerically.

[0062] By placing identification targets whose positions in the absolute coordinate system are known at several locations in the field shown in Figure 1, it is possible to determine the exterior orientation parameters of the camera in the absolute coordinate system.

[0063] In the absolute orientation described above, a scale is assigned to the three-dimensional model obtained by relative orientation. In this case, L1 and L2 in Figure 1 are used, but depending on the elevation of the ridge 200, for example, the distance between identification markers 101 and 102 may be shorter than the set value. In this case, the scale in the X direction is used, which is on the same horizontal plane as identification markers 102 and 103. This problem does not occur in the Y direction. However, if the height and inclination angle of the ridges are set in advance, then simply enter that length. Alternatively, if the goal is not to precisely determine the plant height, and relative changes and approximate values ​​are acceptable, enter the sheet value or an estimated value for the ridge angle as the distance between markers.

[0064] After step S103, the process proceeds to step S104. In step S104, feature points are extracted from the images captured from each camera position. Various techniques for extracting feature points are known.

[0065] Next, the correspondence of feature points between the plurality of images that form the stereo image is specified (feature point matching) (step S105). Various techniques are known for specifying the correspondence of feature points between the plurality of images.

[0066] In identifying the correspondence between the above feature points, identification markers that appear in common across multiple target images are used. The identification markers are identified in the images, and their positions within the screen are also detected. Therefore, they can be used to identify the correspondence between multiple target images.

[0067] Next, the three-dimensional coordinates of the feature points whose correspondences between the stereo images have been identified are calculated (step S106). In step S103, the exterior orientation parameters of the camera at each camera position have been determined. Therefore, the positions of the feature points can be calculated according to the principle of stereo photography measurement (forward intersection method).

[0068] Figure 5 is a diagram showing the principle of the forward intersection method. In Figure 5, the exterior orientation parameters of camera 1 and camera 2 are known, point P is a feature point, p is the screen position of feature point P on the shooting screen of camera 1, and q is the screen position of feature point P on the shooting screen of camera 2.

[0069] The coordinate system used is a local coordinate system set in the field of Fig. 1. An example of this local coordinate system is an XYZ coordinate system with the identification marker 113 as the origin. Of course, an absolute coordinate system can also be used.

[0070] Here, a direction line connecting the position of camera 1 (projection origin) and point p, and a direction line connecting the position of camera 2 (projection origin) and point q are set, and their intersection points are found. The position of this intersection point is point P. The positions and orientations of cameras 1 and 2 in the local coordinate system are known, having been found in step S103. Therefore, by setting a direction line connecting the projection origin of camera 1 and point p, and a direction line connecting the projection origin of camera 2 and point q, and finding the positions of their intersection points, the position (coordinates) of feature point P in the coordinate system can be calculated.

[0071] The above is the principle for determining the coordinates of feature point P. This calculation is performed for all feature points for which correspondences have been identified. The above processing is performed in step S106. A collection of feature points whose positions have been determined is called point cloud data. Point cloud data represents an object as a collection of points whose three-dimensional positions have been determined.

[0072] Next, adjustment calculations are performed for all the camera positions and all the feature points whose positions have been found (step S106). In this process, each parameter is finely adjusted so that the error is minimized.

[0073] For example, in the process of Fig. 3, the exterior orientation parameters of the camera are calculated first, and then the positions of each feature point are calculated based on these exterior orientation parameters. Here, the calculated values ​​of the exterior orientation parameters of the camera contain errors, and these errors cause errors in the calculated values ​​of the positions of the feature points.

[0074] This adjustment calculation is called bundle adjustment calculation, and in order to minimize the above error, calculations are repeated with each parameter value slightly shifted to search for the conditions under which the overall calculation value converges. This calculation minimizes the error in the exterior orientation parameters of the camera, and also minimizes the error in each feature point.

[0075] Next, a three-dimensional model of the subject is created using the optimized feature points. In this example, a three-dimensional model of the plant 300 in FIG. 1 is created. The three-dimensional model is created based on the point cloud data obtained in step S106. An example of a three-dimensional model is three-dimensional computer graphics made up of polygons.

[0076] While this example describes stereo photography using two cameras, calculations can be performed using the same principles even when two or more cameras are installed as multi-view stereo and the number of cameras is expanded. The order of camera position estimation, feature point extraction, and point position calculation in steps S103 to S106 may be changed. Camera position estimation may also be performed using feature points. The adjustment calculation in step S107 may be repeated at a certain distance, and the processes in steps S103 to S107 may be repeated or performed all at once at the end. The manner in which these processes are performed is not limited to this example.

[0077] For example, the process shown in Figure 2 is performed once a day, and a three-dimensional model of the plant is obtained for each day.

[0078] (superiority) The agricultural mulch film 100 is equipped with targets used for three-dimensional photo measurement, so when the agricultural mulch film 100 is laid, the targets are installed at the same time. This reduces the effort required to place targets separately. Furthermore, because the positional relationship between the targets is known, the effort required for positioning targets separately can be reduced.

[0079] 2. Second embodiment One method for making it easier to identify correspondences between multiple images (matching between images) is to use a random dot pattern. FIG. 6 is a top view of the agricultural mulch film of this embodiment installed in a farm field. In this embodiment, a random dot pattern 401 is applied to the surface of the agricultural mulch film 400. The random dot pattern 401 is printed, for example, on the surface of the agricultural mulch film 400. Other than the random dot pattern 401, the agricultural mulch film 400 is the same as the agricultural mulch film 100.

[0080] A random dot pattern is one in which dots are arranged randomly. In this example, the dots are black (or colored) and opaque, while the rest of the area is translucent. Conversely, the dots may be translucent and the rest of the area may be opaque. It is also possible to make the entire area opaque, with the dots being a different color from their surroundings so that they can be distinguished.

[0081] By adjusting the transmittance and occupied area of ​​the random dot pattern, it is possible to adjust the sunlight transmittance of the agricultural mulch film 400. In addition, it is also possible to adjust the sunlight absorption efficiency by setting the color of the random dot pattern.

[0082] 3. Third embodiment It is also possible for the same identification marker to be present repeatedly. In this case, the arrangement of the identification markers is devised so that no two identical identification markers appear in the stereo photographic images. For example, the arrangement pattern can be changed every 5 m so that a set (group) of the same type of identification marker appears on the agricultural mulch film. In this case, the difference in the arrangement pattern allows for identification of each individual crop.

[0083] 4. Fourth Embodiment In the case of Figure 1, the correspondence between the holes (e.g., symbols 121 to 123) that are the growth positions of each individual plant and the identification markers is such that the positional relationship of the four identification markers surrounding the growth position of each individual plant is known, but this is not limited to this, and it is possible to have the positional relationship of one or more identification markers known for the growth position of each individual plant.

[0084] For example, two identification markers may be arranged on either side of each of the holes 121 to 123, or six identification markers may be arranged around each of the holes 121 to 123, or the like.

[0085] 5. Fifth Embodiment Another embodiment is shown in Figure 7. Figure 7 shows an agricultural mulch film 600. Figure 7(A) shows a state in which a ridge 700 is covered from above with the agricultural mulch film 600, as viewed vertically from above. Figure 7(B) is a cross-sectional view of the ridge 700 covered with the agricultural mulch film 600, as viewed in the extension direction of the ridge 700 (Y-axis direction). Note that Figure 7(A) shows only a portion of the longitudinal direction of the ridge 700, and in reality, a similar configuration exists extending in the Y-axis direction.

[0086] FIG. 7 shows a case where crops are planted in two rows along the extension direction of the ridges, with the rows being staggered.

[0087] The crops in the first row are planted in the positions of holes 611-615 formed in the agricultural mulch film 600. The crops in the second row are planted in the positions of holes 621-624 formed in the agricultural mulch film 600. The positions of the holes 611-615 and the holes 621-624 are staggered in the Y-axis direction, ensuring the spacing between the crops in the X-axis direction.

[0088] On the upper surface of the agricultural mulch film 600, a row of identification markers 651 to 654, a row of identification markers 661 to 664, a row of identification markers 671 to 675, and a row of identification markers 681 to 685 are provided. The row of identification markers 651 to 654 and the row of identification markers 671 to 675 are arranged at positions on the upper surface of the ridge 700. The row of identification markers 651 to 654 and the row of identification markers 681 to 685 are arranged at positions on the slopes of the ridge 700. The positional relationship of each identification marker is acquired in advance as known information.

[0089] Furthermore, the group of identification markers 651-654 and 661-664 is offset in position in the Y-axis direction from the group of identification markers 671-675 and 681-685. This is to prevent the identification markers from being hidden as much as possible by the growth of the planted crops.

[0090] As shown by reference numerals 691 and 692, both sides of the agricultural mulch film 600 are covered with soil from above, as in the case of Fig. 1. In addition, an identification marker 631 is placed at a high position on a stake 630.

[0091] In this example, for example, the crop planted at position 612 is identified by identification markers 661 and 662. Also, for example, the crop planted at position 624 is identified by identification markers 674 and 675.

[0092] Crops can also be arranged in three or more rows along the extension direction of the agricultural mulch film 600. Identification markers can also be placed between rows of crops. In Figure 7, crops are planted alternately in the first and second rows, but they can also be planted side by side in the X-axis direction.

[0093] It is also possible to plant crops on the slopes of the ridge 700. For example, suppose crops are planted in four rows along the Y-axis direction in Figure 7. In this case, the two central rows are arranged on the raised top of the ridge 600, and the rows on the right and left sides are arranged on the slopes on both sides of the ridge 600.

[0094] 6.Other The ridges for which the agricultural sheet of the present invention is used are not limited to those having a long, narrow longitudinal shape. For example, the present invention can also be used in an agricultural sheet for covering rectangular ridges. [Explanation of symbols]

[0095] 100...Agricultural mulch film, 101-116...Identification markers, 121-123...Holes formed at positions where crops will be planted, 131, 132...Covered soil, 200...Furrow, 210...Stake, 211...Identification marker, 300...Plants to be grown, 400...Agricultural mulch film, 401...Random dot pattern, 501, 503...Cart, 502, 504...Support, 505...Beam, 506-5 09...camera, 600...agricultural mulch film, 611-615...holes formed in the positions where crops will be planted, 621-624...holes formed in the positions where crops will be planted, 630...stake, 631...identification marker, 651-654...identification marker, 661-664...identification marker, 671-675...identification marker, 681-685...identification marker, 691,692...covered soil, 700...ridges.

Claims

[Claim 1] A camera orientation method using an agricultural sheet covering ground on which plants are planted, comprising: The agricultural sheet is A sheet-like member; a plurality of image recognition markers arranged on the surface of the sheet-like member and used for orienting the camera; Equipped with The ground has upwardly raised ridges formed thereon, The ridges are covered with the agricultural sheet, The agricultural sheet in a state where the ridges are covered is A high portion which is a portion of the upper surface of the ridge; a lower portion located lower than the higher portion and covered with soil; a slope portion that is a slope between the high portion and the low portion; and one of the plurality of image recognition markers is disposed in the high-position portion; another of the plurality of image recognition markers is disposed on the slope portion, A camera orientation method using an agricultural sheet, in which the camera is orientated using an image recognition marker placed in the high portion and an image recognition marker placed in the slope portion.

Images