Landmark, position estimation method, and program

Abstract

To provide a landmark capable of improving the accuracy of position estimation, a position estimation method, and a program.SOLUTION: A landmark 3 installed in a path through which a moving body passes is provided with an extension part 12 as a columnar or conical three dimensional shape part, a code part 20 is provided on a surface of the three dimensional shape part, the code part 20 has a plurality of divided regions in a circumferential direction of the three dimensional shape part, and a code pattern is expressed by changing reflection intensity in units of divided regions. The divided regions are, for example, either high reflection intensity regions or low reflection intensity regions, and a binary code pattern is expressed by an arrangement pattern using the high reflection intensity regions and the low reflection intensity regions.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a landmark, a position estimation method, and a program. [Background technology]

[0002] In autonomous driving, an autonomous vehicle needs to understand the surrounding situation in order to estimate its own position. For this reason, autonomous driving uses a technology to estimate its own position using a sensor called Lidar (Light Detection and Ranging) (see, for example, Patent Document 1). When using this technology, an autonomous vehicle is equipped with Lidar and recognizes the surrounding situation from data acquired by Lidar. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2023-038937 Summary of the Invention [Problem to be solved by the invention]

[0004] It is known that in environments with little change in the shape of the surroundings, such as tunnels, or environments covered with objects that absorb Lidar lasers (for example, snow), the accuracy of self-localization decreases or estimation is impossible. In such environments, road signs (including delineators and snow poles) are one of the few objects that can be easily used for self-localization.

[0005] However, because road signs are generally of the same specifications, it is difficult to distinguish them individually based on their shape or color. Furthermore, roads are prone to occlusion (a phenomenon in which an object is hidden from the view of an autonomous vehicle) depending on traffic conditions, making it impossible to pass through all road signs. In such a situation, using road signs as landmarks for self-localization may result in unintentional confusion with road signs that are located in front of or behind the landmarks, potentially increasing the error in the localization.

[0006] From this perspective, the present invention provides a landmark, a position estimation method, and a program that can improve the accuracy of position estimation. [Means for solving the problem]

[0007] The landmark according to the present invention is installed on a route traveled by a moving object. This landmark has a columnar or conical three-dimensional shape, and a code portion is provided on the surface of the three-dimensional shape. The code portion has a plurality of divided regions in the circumferential direction of the three-dimensional shape, and a code pattern is expressed by changing the reflection intensity for each divided region.

[0008] In the landmarks according to the present invention, the identification information and position information of the landmarks can be given as a code pattern, and therefore, by reading this code pattern, the user's own position can be recognized with high accuracy.

[0009] The divided regions may be either high reflection intensity regions or low reflection intensity regions, and a binary code pattern may be expressed by an arrangement pattern using the high reflection intensity regions and the low reflection intensity regions.

[0010] The code portion may have either a divided area to which one or both of a character and a symbol is added, or a divided area whose width is changed. The code section may further include a reference code section for specifying a reading range of the code pattern.

[0011] In this way, even if the landmark is rotated, misreading of the code pattern can be suppressed.

[0012] The three-dimensional shape portion may further include an end emphasis portion provided on one or both sides of the code portion in the axial direction.

[0013] This makes it easier to distinguish between foreign objects (such as signs and retroreflective materials worn by people) and landmarks, and also makes it easier to make landmarks comply with existing road standards.

[0014] A position estimation method according to the present invention is a method for estimating the position of a moving object using the landmarks described above. This position estimation method includes a point cloud data acquisition step, a landmark detection step, a code pattern reading step, and a position identification step. In the point cloud data acquisition step, point cloud data is acquired using a detection unit possessed by the moving object. In the landmark detection step, a point cloud corresponding to the landmark is detected from the point cloud data. In the code pattern reading step, the code pattern is read from the point cloud recognized as the landmark. In the position identification step, a position is identified based on the read code pattern.

[0015] The position estimation method according to the present invention can read the code pattern of a landmark and obtain the landmark's identification information and position information from the read code pattern, thereby enabling the user to accurately recognize their own position.

[0016] In the code pattern reading step, it is preferable to read the code pattern when the reading angle is equal to or greater than a code display angle that indicates the angle occupied by one of the code patterns.

[0017] A program according to the present invention causes a computer to execute the above-described position estimation method. The program according to the present invention can read the code pattern of a landmark and obtain the landmark's identification information and position information from the read code pattern, thereby enabling the device to recognize its own position with high accuracy. [Effects of the Invention]

[0018] According to the present invention, it is possible to improve the accuracy of position estimation. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a schematic configuration diagram of an autonomous mobile system including a position estimation system according to an embodiment of the present invention. [Figure 2] FIG. 1 is a schematic diagram of a moving body. [Figure 3] This is an image of the detection unit (multi-layer Lidar) equipped on a moving object. [Figure 4] FIG. 1 is a schematic diagram illustrating the configuration of landmarks for position estimation. [Figure 5] FIG. 10 is an enlarged view of the main part of the landmark with the symbol portion enlarged. [Figure 6] 10 is an example of a pattern of arrangement of high reflection intensity regions and low reflection intensity regions in a code portion. [Figure 7] 10A and 10B are diagrams for explaining the surface division of the code part, in which (a) shows the code part when the number of surface divisions is "2", and (b) shows the code part when the number of surface divisions is "3". [Figure 8] This is an example of a code pattern when the number of surface divisions is "2". [Figure 9] This is an example of a case where the rotation angle of the sign part is smaller than the sign display angle, where (a) shows the state before rotation, (b) shows the state where the sign part has rotated clockwise by "60°" from the state in (a), and (c) shows the state where the sign part has rotated clockwise by "120°" from the state in (a). [Figure 10] 10A and 10B are diagrams for explaining misreading of code patterns, in which (a) is the code pattern for the decimal digit "1" and (b) is the code pattern for the decimal digit "4." [Figure 11] This is an example of a base code part. (a) shows a code pattern with a display digit count of "3" that can be used as a base code part, and (b) and (c) show cases where symbols or special patterns are added. [Figure 12] 10A and 10B are diagrams illustrating the case where a code pattern is read using a reference code portion, where (a) shows the case where the digit "1" is read, and (b) shows the case where the digit "4" is read. [Figure 13] 10A and 10B are diagrams for explaining a method for avoiding misreading of code patterns, in which (a) is the code pattern for the decimal digit "1" and (b) is the code pattern for the decimal digit "4." [Figure 14] 10A and 10B are diagrams for explaining the problems and methods for avoiding the problems that occur when multiple code parts are written in the same way, where (a) shows a case where the boundaries of each code part are connected and the range becomes unclear, (b) shows a case where each code part is identified by its length in the axial direction, and (c) shows a case where each code part is identified by being shifted in the circumferential direction. [Figure 15] 10 is a graph showing the reflection intensity for each material at each angle. [Figure 16] 10 is a graph showing the reflection intensity for each material at different distances. [Figure 17] FIG. 1 is a diagram for explaining the inter-layer distance of a multi-layer Lidar. [Figure 18] 1 is an example of a flowchart illustrating the overall processing of a position estimation method in a position estimation system according to an embodiment. [Figure 19] 10 is a flowchart illustrating an example of a road sign detection process. [Figure 20] 10 is an image diagram of the segmentation process, where (a) shows the state before processing and (b) shows the state after processing. [Figure 21] 10A and 10B are conceptual diagrams of width-based filtering processing, where (a) is a diagram for explaining the width of landmarks, and (b) and (c) are examples of segmented point clouds. [Figure 22]10A and 10B are conceptual diagrams of a filtering process based on length, in which (a) is a diagram for explaining the length of the code portion, and (b) and (c) are examples of segmented point clouds. [Figure 23] This is an illustration of landmark detection. (a) and (b) are processing images when the landmark is outside the detection distance range, and (c) is a processing image when the landmark is within the detection distance range. [Figure 24] This is an image diagram of the length-based filtering process using tracking points. (a) shows the relationship between the detection unit and landmarks at time "0", and (b) shows the relationship between the detection unit and landmarks at time "t", which is "t" ahead of time "0". [Figure 25] 1 is a diagram illustrating the positional relationship between landmarks and a detection unit. [Figure 26] This is an example of a landmark. [Figure 27] This is a graph showing the positions of recorded points at the "-15000mm" point. [Figure 28] This is a graph showing the positions of the recorded points when the position reaches the "-5000mm" point. [Figure 29] This is a graph showing the positions of the recorded points when the robot has moved to the "15000mm" point. [Figure 30] Examples of landmarks: (a) is a symbol representing the number "5" with a surface division number of "2", (b) is a symbol representing the number "2" with a surface division number of "2", and (b) is a symbol representing the number "6" with a surface division number of "2". [Figure 31] This is the result of detecting the code portion representing the number "5." [Figure 32] This is the result of detecting the code portion representing the number "2." [Figure 33] This is the result of detecting the code portion representing the number "6." [Figure 34] This is an image of an autonomous mobile system being used inside a tunnel. [Figure 35]This is an example of how the autonomous mobile system would be used inside a warehouse. [Figure 36] 10A and 10B are diagrams illustrating variations in the number of digits displayed in the sign portion. [Figure 37] FIG. 10 is a diagram for explaining variations in the encoding section. [Figure 38] 10A and 10B are schematic diagrams of a landmark according to a first modified example, in which FIG. 10A shows the landmark in a disassembled state, and FIG. 10B shows the landmark in a state in use. [Figure 39] FIG. 10 is a schematic configuration diagram of a landmark according to a second modified example. [Figure 40] 10A and 10B are schematic diagrams of a landmark according to a third modified example, in which (a) shows the landmark in a disassembled state, and (b) shows the landmark in a state in use. [Figure 41] 10A and 10B are schematic diagrams of a landmark according to a fourth modified example, in which (a) shows the landmark in a disassembled state, and (b) shows the landmark in a state in use. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Each drawing is merely a schematic illustration to allow a sufficient understanding of the present invention. Therefore, the present invention is not limited to the illustrated examples. In each drawing, common or similar components are designated by the same reference numerals, and redundant explanations thereof will be omitted.

[0021] <Configuration of the position estimation system according to the embodiment> The configuration of a position estimation system 1A according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a schematic configuration diagram of an autonomous mobile system 1 including the position estimation system 1A. The position estimation system 1A may constitute a part of another system, and in this embodiment, it is assumed that the position estimation system 1A constitutes a part of an autonomous mobile system 1 in which a mobile object 2 moves autonomously without being operated. In other words, the mobile object 2 estimates its own position using the position estimation system 1A and moves autonomously based on the estimated position.

[0022] In the autonomous mobile system 1 shown in Fig. 1, a mobile object 2 (for example, a self-driving car) moves autonomously. In this embodiment, it is assumed that a starting point, a destination point, and a route from the starting point to the destination point are specified for the mobile object 2, and the mobile object 2 moves along the determined route without being controlled. Note that the mobile object 2 may also calculate the destination point and the route to the destination point based on information provided by other systems or devices or information detected by the mobile object 2 itself, and the mobile object 2 may move along the calculated destination point and route.

[0023] The autonomous mobile system 1 shown in FIG. 1 mainly comprises a mobile object 2 and landmarks 3 for position estimation. The landmarks 3 for position estimation are, for example, road signs (including delineators and snow poles) and are installed on the route traveled by the mobile object 2. As will be described in detail later, each landmark 3 is set with information that allows it to be individually identified, and this information is attached to the landmark 3 so that it can be read by a detection unit 5 (for example, Lidar) included in the mobile object 2. The mobile object 2 estimates its own position by detecting the landmarks 3.

[0024] 2 shows a schematic configuration diagram of the mobile object 2. The mobile object 2 includes a moving unit 4, a detecting unit 5, a communicating unit 6, a storage unit 7, and a control unit 8. The moving unit 4 is a means for moving the moving body 2, and the movement here includes walking or running on the ground, sailing on water, flying in the air, etc. In this embodiment, moving on the ground is assumed, and the moving unit 4 is composed of, for example, wheels, a power source that rotates the wheels, and a transmission mechanism (for example, gears and shafts) that transmits the power generated by the power source to the wheels.

[0025] The detection unit 5 is a means for detecting the situation around the moving object 2. The detection unit 5 is, for example, a multi-layer Lidar, and has an irradiation unit that irradiates laser light and a receiving unit that receives the laser light reflected from the target. In a multi-layer Lidar, the irradiation unit emits laser light at specified angular intervals of about 1 to 2 degrees in the vertical direction and at dense intervals of 0.1 to 0.4 degrees in the horizontal direction while rotating using a motor (see Figure 3). Figure 3 is an image diagram of a multi-layer Lidar. This allows the detection unit 5 to have a planar detection range and to acquire the shape of the target as point cloud data. The detection unit 5 has a function of measuring the reflection intensity of the received laser light.

[0026] The communication unit 6 is a device that realizes communication with other devices or satellites, and is configured, for example, with an antenna and its peripheral devices. The mobile object 2 can communicate with, for example, a management device (not shown) that manages the entire autonomous mobile system 1 via the communication unit 6. The mobile object 2 and the management device do not need to be in a state where they can communicate at all times; for example, the mobile object 2 may communicate with the management device when it moves into an area where communication is possible. Furthermore, the mobile object 2 can receive radio waves from a global navigation satellite system (GNSS) via the communication unit 6. The mobile object 2 may estimate its position using landmarks 3, for example, in an environment where it cannot receive radio waves from the global navigation satellite system, when the accuracy of position estimation based on the global navigation satellite system is low, or when it enters a predetermined specific area.

[0027] The storage unit 7 is composed of storage media such as a RAM (Random Access Memory), a ROM (Read Only Memory), an HDD (Hard Disk Drive), an SSD (Solid State Drive), and a flash memory. The storage unit 7 stores information for autonomously moving the mobile object 2. The information stored in the storage unit 7 includes information for estimating the position of the mobile object 2. Note that some or all of the information stored in the storage unit 7 may be stored in a location other than the mobile object 2 (for example, a management device) and may be acquired as needed via the communication unit 6.

[0028] The memory unit 7 stores reflection intensity information relating to the reflection intensity of the components that make up the landmark 3. This reflection intensity information corresponds to, for example, the wavelength of the laser light emitted by the detection unit 5, and includes information about the reflection intensity of a high reflection intensity region with high reflection intensity and a low reflection intensity region with low reflection intensity, which will be described later. The reflection intensity information may be information that has been pre-mapped, which includes information about the distance from the moving object 2 to the landmark 3 and information about the reflection intensity. The memory unit 7 also stores landmark position information relating to the position where the landmark 3 is provided. This landmark position information is, for example, information about the coordinate values of the landmark 3 and the conditions under which the landmark 3 is set.

[0029] The control unit 8 is, for example, a CPU (Central Processing Unit) and peripheral electronic devices, and executes program processing. The control unit 8 realizes various functions for autonomously moving the moving body 2 by executing the program. The program executed by the control unit 8 can be stored in a computer-readable recording medium (e.g., CD-ROM) and provided. The program can also be provided via a network such as the Internet. Here, the process of estimating the position of the moving body 2 using the landmarks 3 will be particularly described, and a detailed description of the overall process related to autonomous movement will be omitted. The process related to autonomous movement is not particularly limited, and may be realized by, for example, general processing.

[0030] The landmark 3 for position estimation will be described with reference to Fig. 4. In this embodiment, the landmark 3 will be described assuming a two-stage retractable snow pole. A snow pole is a type of road sign installed on the side of the road, and is a safety device that helps drivers recognize the road shape. Snow poles are installed in cold regions and areas with heavy snowfall.

[0031] The landmark 3 shown in FIG. 4 has a support portion 11 and an extension portion 12. The extension portion 12 is slidable relative to the support portion 11 and is used in an extended state when there is snow. The support portion 11 and the extension portion 12 are an example of a columnar three-dimensional portion. A code portion 20 is provided on the surface of one or both of the support portion 11 and the extension portion 12. The code portion 20 is divided into two or more divisions in the circumferential direction of the three-dimensional portion, thereby having a plurality of divided areas, and the reflection intensity is changed for each divided area. Note that the landmark 3 may be any mark that can be used for position estimation, and the configuration of the landmark 3 is not particularly limited. The landmark 3 may have only the support portion 11 without the extension portion 12, or may have only the extension portion 12 without the support portion 11.

[0032] The code portion 20 will be described with reference to Fig. 5. Fig. 5 is an enlarged view of the code portion 20 of the landmark 3. In this embodiment, it is assumed that a high reflection intensity region and a low reflection intensity region are provided for the landmark 3, and a code pattern is assigned to the landmark 3 according to each pattern. The landmark 3 shown in Fig. 5 has the code portion 20 that expresses the code pattern and an edge emphasis portion 30 that emphasizes the boundary of the code portion 20.

[0033] The edge enhancer 30 serves to distinguish between foreign objects (such as signs and retroreflective materials worn by people) and landmarks 3. Furthermore, if the area of a high reflection intensity band of a road sign is specified by road standards, and the high reflection intensity region in the code section 20 does not conform to the road standards, the width of the edge enhancer 30 can be adjusted to conform to the existing road standards while still allowing the present invention to be adopted. For example, if the area of the high reflection intensity band is insufficient to meet the road standards, the edge enhancer 30 can be lengthened to conform to the road standards. Depending on the road standards, road conditions, and installation environment, the edge enhancer 30 is not necessarily required (i.e., the landmark 3 does not need to have the edge enhancer 30). Examples of cases where the edge enhancer 30 is not required include warehouses and construction sites. In these cases, the general public is not allowed to enter, and the reflection intensity of objects other than the landmark 3 used for position estimation can be managed, making it possible to distinguish between landmarks 3 and foreign objects without using the edge enhancer 30.

[0034] The end emphasis portion 30 is provided on one or both sides of the symbol portion 20 in the axial direction of the support portion 11 or the extension portion 12 (three-dimensional shape portion). In the landmark 3 shown in Figure 5, a first end emphasis portion 31 is provided above the symbol portion 20, and a second end emphasis portion 32 is provided below the symbol portion 20. The end emphasis portion 31 indicates the upper end of the symbol reading, and the end emphasis portion 32 indicates the lower end of the symbol reading.

[0035] The code section 20 has a plurality of divided regions in the circumferential direction of the support section 11 or the extension section 12 (three-dimensional shape section), and forms high reflection intensity regions or low reflection intensity regions in divided region units. As a result, the code section 20 expresses binary information by the arrangement pattern of the high reflection intensity regions and low reflection intensity regions, where the high reflection intensity regions are represented as "1" and the low reflection intensity regions as "0" (or vice versa). The number of digits displayed for the information expressed by the code section 20 can be changed by changing the number of divided digits.

[0036] Fig. 6 shows an example of the arrangement pattern of high reflection intensity regions and low reflection intensity regions in the code section 20. Fig. 6 shows the case where the number of display digits is "3" (where there are three divided regions), and decimal numbers "0" to "7" are represented by the arrangement pattern of the high reflection intensity regions and low reflection intensity regions. For example, if all divided regions are low reflection intensity regions, it represents "0," if the divided region on the right is a high reflection intensity region and the others are low reflection intensity regions, it represents "1," and if all divided regions are high reflection intensity regions, it represents "7."

[0037] If the landmark 3 is a road signpost and the road signpost is placed on the side of the road, it is difficult to detect the entire surface of the landmark 3 from the moving body 2. In other words, when the moving body 2 is moving straight, the range irradiated with the laser light from the detection unit 5 is the side facing the moving body 2, and the laser light is not irradiated on the side not facing the moving body 2. For this reason, an arrangement pattern of high reflection intensity regions and low reflection intensity regions in the code unit 20 is set in the range irradiated with the laser light from the detection unit 5 (for example, 45 to 270°).

[0038] Note that snow poles installed over fixed road signs and the extensions (extending from the support) of two-stage retractable snow poles may not be completely fixed and may be designed to allow horizontal rotation. In such cases, the range illuminated by the laser light from the detection unit 5 changes depending on the rotation of the snow pole. In this case, it is advisable to divide the periphery of the landmark 3 into multiple planes and apply the same code pattern to each plane. When a horizontal cross section of the code unit 20 is displayed, the angle occupied by one code pattern is referred to as the "code display angle." When the number of plane divisions is "2," the code display angle is "360° / 2 = 180°," and when the number of plane divisions is "3," the code display angle is "360° / 3 = 120°" (see Figure 7). Figure 7 is a diagram for explaining the plane division of the code unit 20. (a) shows the code unit 20 when the number of plane divisions is "2," and (b) shows the code unit 20 when the number of plane divisions is "3."

[0039] An example of a code pattern when the number of surface divisions is "2" is shown in Figure 8. In Figure 8(a), the code pattern of the number "6" is attached to both the first and second surfaces, in Figure 8(b), the code pattern of the number "5" is attached to both the first and second surfaces, and in Figure 8(c), the code pattern of the number "3" is attached to both the first and second surfaces. In this way, the same code pattern is attached to each surface. As a result, even if the landmark 3 is rotated "180°", the code pattern is included in the range irradiated by the laser light of the detection unit 5.

[0040] Here, it is acceptable if the landmark 3 is rotated by the code display angle, but if the rotation angle of the landmark 3 is smaller than the code display angle or if the rotation angle of the landmark 3 is larger than the code display angle, misreading of the code pattern may occur. Referring to FIG. 9, an example of a case where the rotation angle of the landmark 3 is smaller than the code display angle is shown. FIG. 9(a) shows the state before rotation, FIG. 9(b) shows the state where the code section 20 has rotated by "60°" clockwise from the state in (a), and FIG. 9(c) shows the state where the code section 20 has rotated by "120°" clockwise from the state in (a).

[0041] With reference to FIG. 10 , misreading of the code pattern will be explained in more detail. Here, we assume that the decimal digits "1" and "4" are represented by a display digit count of "3" and a surface division count of "2." When the code pattern for the decimal digit "1" shown in the middle of FIG. 10( a) is rotated by "120°," the code pattern for the digit "4" is displayed on the surface facing the moving object 2. As such, if the code unit 20 is a standalone unit, the rotation of the code unit 20 may result in misreading of the code pattern. Assuming the display digit count is "3," the code pattern may be misread within the group of digits "1, 2, 4" surrounded by the dashed line and the group of digits "3, 5, 6" surrounded by the dashed line shown in FIG. 6. Therefore, if the code unit 20 rotates because it is not directly fixed to the ground, for example, some measure is required to avoid misreading of the code pattern due to the rotation of the code unit 20.

[0042] In this embodiment, a reference code section 21 is provided as a method for avoiding misreading of the code pattern due to rotation of the code section 20 (see FIG. 5). The reference code section 21 shown in FIG. 5 is used to limit the reading range of the second code section 22 and the third code section 23 (which may be either or both of the reading start and end positions of the code pattern). A code pattern that does not produce identical patterns until it completely overlaps even when rotated is set as the reference code section 21, and the ranges of the second code section 22 and the third code section 23 are determined based on the position of the reference code section 21. Note that code patterns whose starting positions are unknown cannot be used as the reference code section 21 (for example, the code patterns of the numbers "0" and "7" in FIG. 6).

[0043] Fig. 11 shows an example of the base code section 21. Fig. 11(a) shows a pattern that can be used as the base code section 21 among code patterns for the number of display digits "3". When the number of display digits is "3", there are two patterns that can be used as the base code section 21: a group of numbers "1, 2, 4" surrounded by a dashed line shown in Fig. 6, and a group of numbers "3, 5, 6" surrounded by a dashed line shown in Fig. 6. Note that it is also possible to add symbols and special patterns to the extent that they do not interfere with the determination of the ranges of the second code section 22 and the third code section 23 (see Figs. 11(b) and (c)).

[0044] 12 shows a case where the reading range of the second code portion 22 is limited using the group of numbers "3, 5, 6" in FIG. 6 as the reference code portion 21. FIG. 12(a) shows a case where the reference code portion 21 reads the number "1" displayed in the second code portion 22, and FIG. 12(b) shows a case where the reference code portion 21 reads the number "4" displayed in the second code portion 22. As shown in FIG. 12, by using the reference code portion 21, it is possible to limit the reading range of the second code portion 22.

[0045] As a method for preventing misreading of a code pattern due to rotation of the code section 20, the provision of a reference code section 21 has been described. Alternatively, misreading of a code pattern can be prevented by changing the width of the divided regions of the code section 20. For example, the width of the divided regions can be set so that each digit has a different width. Specifically, as shown in FIG. 13, the width of the divided region for the rightmost digit is set to be the narrowest, the width of the divided region for the second digit from the right is set to be wider than that of the rightmost digit, and the width of the divided region for the third digit from the right (leftmost digit) is set to be the widest. FIG. 13 is a diagram for explaining a method for preventing misreading of a code pattern. (a) shows the code pattern for the decimal digit "1," and (b) shows the code pattern for the decimal digit "4." In this way, patterns that appear identical in FIG. 10 can be distinguished.

[0046] 14(a), when the reference numerals 21, 22, and 23 are written in the same way, the boundaries and ends of the reference numerals 21, 22, and 23 are connected, making the ranges of the reference numerals 21, 22, and 23 unclear. As a method for avoiding this unclear range of the reference numerals 21, 22, and 23, it is effective to identify the reference numerals 21, 22, and 23 by their lengths in the axial direction, or to displace the reference numerals 21, 22, and 23 in the circumferential direction, as shown in FIG.

[0047] We will explain the materials for high reflectivity areas and low reflectivity areas. For the material of high reflectivity areas, it is best to use retroreflective material. Retroreflective material is a material that has the property of reflecting light in the direction of the light when it is irradiated with light. Due to its characteristics, it is known that when retroreflective material comes within the scanning range of a Lidar that can detect reflection intensity, only the area of the retroreflective material will have high reflection intensity.

[0048] The material for the low reflection intensity area is not particularly specified, but it is desirable that it be a material that reflects little when light is shone on it. In order to improve the ability to distinguish it from general road sign materials (such as vinyl chloride) or retroreflective materials, it is advisable to use a low reflection material for the low reflection intensity area, for example. In particular, using an ultra-low reflection material is effective because it can reduce the failure rate of distinguishing between high reflection intensity and low reflection intensity areas. Figure 15 shows the reflection intensity for each material by angle. Figure 16 also shows the reflection intensity for each material by distance.

[0049] The size of the encoding unit 20 will be described with reference to FIG. 17. FIG. 17 shows the inter-layer distance of a multi-layer Lidar with a vertical angle of 2°. The unit of distance shown in FIG. 17 is (mm). Performing calculations on all point cloud information acquired by Lidar would take an enormous amount of calculation time, making it unsuitable for control that requires real-time processing, such as for self-driving cars. Therefore, it is important to cut out unnecessary information as much as possible. The same is true for reading the encoding unit 20 in the present invention; it is important to detect the position of the encoding unit 20 early (approximately 10 to 20 m ahead) before reading the encoding unit 20.

[0050] Because retroreflective materials are commonly used, it is difficult to distinguish between the code 20 and other objects simply by detecting a single location with high reflection intensity. In this embodiment, the characteristics of multi-layer Lidar are also taken into consideration, and it is assumed that the code 20 can be detected early by acquiring two or more points separated vertically. Using the inter-layer distance shown in Figure 17 as a reference, in order to acquire two or more points separated vertically from a position 15 m away, a high reflection intensity area of at least 53 cm is required.

[0051] If the standard is a multi-layer Lidar with a vertical angle of 1°, a high-reflection-intensity region of 26 cm or more is sufficient. However, since recognizing points that are vertically distant also serves to distinguish from signs, retroreflective materials worn by people, etc., it is more realistic to maintain the vertical length of the high-reflection-intensity region at a distance of 53 cm or more and expand the detection range to approximately 30 m. In other words, while the vertical length of the high-reflection-intensity region can be shortened with a multi-layer Lidar with a vertical angle of 1°, shortening it is not recommended from the perspective of distinguishing from other objects; instead, it is desirable to expand the detection range. This allows for earlier detection of the code portion 20 and an earlier start timing for reading.

[0052] <Regarding the position estimation method by the position estimation system according to the embodiment> The position estimation process by the position estimation system 1A according to the embodiment will be described with reference to Fig. 18 and Fig. 19. Fig. 18 is an example of a flowchart showing the overall process of the position estimation method in the position estimation system 1A according to the embodiment. Fig. 19 is an example of a flowchart showing road sign detection processing. Note that the control shown in Fig. 18 is started, for example, when the moving object 2 starts moving.

[0053] (Overall processing) The control unit 8 of the moving object 2 determines whether the moving object 2 is on a travel route (a road, path, passageway, factory, warehouse, construction site, or other area used by the moving object 2) to determine whether to use the position estimation system 1A (step S1). The process of step S1 is effective because it is necessary to distinguish between road signs and retroreflective materials attached to signs and pedestrian clothing. Note that the process of step S1 is not necessarily required in places such as tunnels and warehouses. If the result of step S1 is "No," the process of step S1 is executed again after a period of time.

[0054] If step S1 returns "Yes," the control unit 8 determines whether the moving object 2 is within an area where a landmark 3 (here, a road sign is assumed) is installed (step S2). In places where it is clear that no landmark 3 is installed, there is no need to detect the landmark 3. Therefore, when using on public roads, it is desirable to set an installation area for the landmark 3 (for example, the area indicated by the symbol AR in FIG. 1) on a map inside the moving object 2 (for example, a car), and detect the landmark 3 when the moving object 2 enters that area. If step S2 returns "No" (including when the moving object 2 leaves the installation area AR), the process returns to step S1.

[0055] The determination in step S1 of whether the moving object 2 is on the travel path and the determination in step S2 of whether the moving object 2 is in the installation area AR are performed using a method such as odometry using GNSS, GPS, an accelerometer, wheel speed, etc. Note that steps S1 and S2 can also be performed simultaneously.

[0056] If the answer is "Yes" in step S2, the detection unit 5 (assumed to be Lidar in this case) emits laser light around the moving object 2, receives the laser light reflected by surrounding objects, and acquires data on the surrounding environment (step S3). In this embodiment, data on the surrounding environment converted into a point cloud (point cloud data) is used.

[0057] Next, the control unit 8 determines whether the "code reading FLAG" is "TRUE" (step S4). The "code reading FLAG" is a flag indicating whether or not the code pattern reading process is possible, and is set to "TRUE" if the code pattern reading process is possible, and is set to "FALSE" if the code pattern reading process is not possible. Since the "code reading FLAG" is set to "FALSE" as the initial value, the initial determination in step S4 is "No." If the determination in step S4 is "No," the control unit 8 executes detection processing of the landmark 3 (here, assumed to be a road sign) using the point cloud data acquired by the detection unit 5 (step S5).

[0058] (Road sign detection processing) The road sign detection process (step S5) will be described with reference to Fig. 19. The control unit 8 reads the point cloud data acquired by the detection unit 5 (step S41), and in subsequent processing performs processing to detect road signs as landmarks 3 included in the point cloud data. The code attached to the landmark 3 cannot be read unless the landmark 3 is extracted from a large number of point clouds. In this embodiment, the landmark 3 is provided with edge enhancement units 31 and 32 as high reflection intensity regions (see Fig. 5), and here, the case where the landmark 3 is detected based on the edge enhancement units 31 and 32 will be described.

[0059] The laser from the detection unit 5 that is emitted onto the edge emphasis units 31, 32 is reflected back to the detection unit 5 with low loss, and is therefore recognized by the detection unit 5 as a point of high reflection intensity. Data from a single point irradiated onto a high reflection intensity area cannot be distinguished from retroreflective material attached to signs, etc. or the clothing of pedestrians, etc. However, by detecting data from two or more points irradiated in the vertical direction onto the two edge emphasis units 31, 32, it becomes possible to distinguish between road signs as landmarks 3 and other objects (i.e., it becomes possible to recognize landmarks 3).

[0060] The control unit 8 determines whether or not a value of high reflection intensity has been detected within the detection range of the detection unit 5 (that is, whether or not the detection unit 5 has detected reflected light with high reflection intensity) (step S42). If the result in step S42 is "No," the "code reading FLAG" remains "FALSE," and the road sign detection process (step S5) ends. If the result in step S42 is "Yes," the control unit 8 performs segmentation of the high reflection intensity points (step S43). An image of the segmentation process is shown in FIG. 20. FIG. 20 is an image diagram of the segmentation process, where (a) shows the state before processing and (b) shows the state after processing. As shown in FIG. 20, the points are segmented into groups of points near high reflection intensity.

[0061] Next, the control unit 8 determines whether the width of the point cloud for each segmented segment is smaller than the width of the landmark 3 (here, assuming a road sign) (step S44). If the result in step S44 is "No", the "code reading FLAG" remains "FALSE", and the road sign detection process (step S5) ends. That is, filtering by width is performed to remove from the processing the point cloud thicker than the target landmark 3 (here, assuming a road sign) from the information of the segmented point cloud.

[0062] An image of the filtering process by width is shown in Fig. 21. Fig. 21 is an image diagram of the filtering process based on width. (a) is a diagram for explaining the width of the landmark 3 (here, assuming a road sign), and (b) and (c) are examples of the segmented point cloud. The width h1 of the point cloud shown in Fig. 21(b) is smaller than the width H of the landmark 3 shown in (a) (h1 < H), so the point cloud is not removed by the process in step S44. On the other hand, the width h2 of the point cloud shown in Fig. 21(c) is larger than the width H of the landmark 3 shown in (a) (h2 > H), so the point cloud is removed by the process in step S44.

[0063] Next, the control unit 8 determines whether there are points separated by "50 cm" or more in the vertical direction within the segmented point cloud (step S45). The "50 cm" in the vertical direction is a value corresponding to the layer distance explained in Fig. 17 and is only an example of the separation distance between the end highlighting portions 31 and 32. If the result in step S45 is "Yes", "TRUE" is set for the "code reading FLAG" (step S50), and the road sign detection process (step S5) ends. That is, filtering by length is performed, and when the vertical length of the segmented point cloud exceeds the separation distance between the end highlighting portions 31 and 32, the point cloud is regarded as part of the landmark 3 (the code portion 20).

[0064] An image of the filtering process based on length is shown in FIG. 22. FIG. 22 is an image diagram of the filtering process based on length, where (a) is a diagram for explaining the length of the code portion 20, and (b) and (c) are examples of segmented point clouds. The length m1 of the point cloud shown in FIG. 22(b) is greater than or equal to the length M of the code portion 20 shown in (a) (m1≧M), so the point cloud is regarded as part of the landmark 3. On the other hand, the length m2 of the point cloud shown in FIG. 22(c) is smaller than the length M of the code portion 20 shown in (a) (m2<M), so the point cloud is not regarded as part of the landmark 3.

[0065] Referring to FIG. 23, the detection process (steps S41 to S45) of the landmark 3 using the end emphasis portions 31 and 32 described so far will be described. FIG. 23 is an image diagram of the detection of the landmark 3, where (a) and (b) are processing images when the landmark 3 is outside the detection distance range, and (c) is a processing image when the landmark 3 is within the detection distance range. As shown in FIG. 23(a), the landmark 3 within the road sign detection distance (for example, about 10 to 30 m) will be irradiated with the laser light of the Lidar on each of the end emphasis portions 31 and 32, and it is determined as "Yes" in step S45 and recognized as a landmark.

[0066] On the other hand, as shown in FIG. 23(b), the landmark 3 outside the road sign detection distance (for example, about 10 to 30 m) is not irradiated with the laser light of the Lidar on any of the end emphasis portions 31 and 32, and it is determined as "No" in step S42 and excluded from the process. Also, as shown in FIG. 23(c), the landmark 3 outside the road sign detection distance (for example, about 10 to 30 m) is irradiated with the laser light of the Lidar on one of the end emphasis portions 31 and 32, and it is determined as "No" in step S45 (that is, there is no point cloud exceeding the separation distance between the end emphasis portions 31 and 32) and not recognized as a landmark.

[0067] If step S45 returns "No" (i.e., there is no point cloud that exceeds the separation distance between the edge emphasis sections 31 and 32), it is assumed that there is actually a landmark 3 within the road sign detection distance (for example, about 10 to 30 m), but that it has not been recognized as a landmark for some reason. For example, this could be because the laser light does not exceed the separation distance between the edge emphasis sections 31 and 32 due to a problem with how the laser light hits, or because occlusion occurs due to an obstacle and the laser light is not irradiated to one of the edge emphasis sections 31 and 32 (or is irradiated but cannot be received), resulting in detection on only one side. Therefore, in steps S46 to S49, processing is performed on points that are within the road sign detection distance (for example, about 10 to 30 m) but have not been recognized as a landmark.

[0068] If the answer is "No" in step S45 (i.e., if there is no point cloud that exceeds the separation distance between the edge emphasis units 31 and 32), the control unit 8 determines whether the "Tracking FLAG" is "TRUE" (step S46). Since the "Tracking FLAG" is set to "FALSE" as its initial value, the initial determination in step S46 is "No." If the answer is "No" in step S46, the positions of the segmented point cloud are recorded (as tracking points), and the "Tracking FLAG" is set to "TRUE" (step S47). Then, the "Code Reading FLAG" is left at "FALSE," and the road sign detection process (step S5) is terminated.

[0069] On the other hand, if the answer is "Yes" in step S46 (if the position of the segmented point cloud has been recorded), the control unit 8 adds the estimated movement difference obtained from the wheel speed, gyro sensor, on-board sensor, etc. to the recorded position (tracking point) to obtain a new tracking point (step S48). This causes the tracking point to be updated over time. Next, the control unit 8 uses the updated tracking point and the newly obtained point cloud to perform the length-based filtering process described in step S45 again (step S49). If the answer is "Yes" in step S49 (if the vertical length of the point cloud exceeds the separation distance between the edge enhancement units 31 and 32), the control unit 8 considers the point cloud to be part of the landmark 3 (here, a road sign is assumed), sets "TRUE" to the "code reading FLAG" (step S50), and ends the road sign detection process (step S5). On the other hand, if "No" in step S49, the point group is not considered to be part of the landmark 3, the "code reading FLAG" remains "FALSE", and the road sign detection process (step S5) ends.

[0070] An image of the filtering process based on length using tracking points, which is realized in steps S46 to S49, is shown in Fig. 24. Fig. 24 is an image diagram of the filtering process based on length using tracking points, in which (a) shows the relationship between the detection unit 5 and landmarks 3 at time "0", and (b) shows the relationship between the detection unit 5 and landmarks 3 at time "t", which is "t" ahead of time "0".

[0071] At time "0" shown in FIG. 24(a), it is assumed that the landmark 3 is present within the road sign detection distance (e.g., approximately 10 to 30 m), but only the edge emphasis portion 31 can be detected, and the edge emphasis portion 32 is not. In this case, in step S47, the position of the edge emphasis portion 31 is stored. Next, at time "t" shown in FIG. 24(b), it is assumed that the distance between the detection unit 5 and the landmark 3 becomes shorter, so that the edge emphasis portion 31 is not detected, and only the edge emphasis portion 32 can be detected. In this case, in step S48, the estimated movement difference at time "t" is added to the position of the edge emphasis portion 31 at time "0," and the position of the edge emphasis portion 31 at time "t" is estimated. Then, in step S49, a length-based filtering process is performed based on the estimated position of the edge emphasis portion 31 at time "t" and the detected position of the edge emphasis portion 32 at time "t," and the presence of the landmark 3 is confirmed.

[0072] When the processing of step S5 is completed, the process returns to step S1, and the processing of steps S1 to S3 is executed (see FIG. 18). Then, the control unit 8 determines whether or not the "code reading FLAG" is "TRUE" (step S4). If there is a point group regarded as landmark 3 by the previous processing of step S5, the "code reading FLAG" is set to "TRUE", and if there is no point group regarded as landmark 3, the "code reading FLAG" is set to "FALSE".

[0073] If the result of step S4 is "Yes" (i.e., "TRUE" is set in the "code reading FLAG"), the control unit 8 acquires the road sign distance to the road sign detected in step S5 (step S6), and determines whether or not "road sign distance < road sign detection range" (step S7). If the road sign distance is equal to or greater than the road sign detection range ("No" in step S7), the "code reading FLAG" is set to "FALSE" (step S8), and the process returns to step S1. For example, if a road sign was detected in step S5, but the road sign distance has become greater (farther) than the road sign detection range due to the movement of the moving object 2, the result of step S7 is "No."

[0074] If the road sign distance is smaller than the road sign detection range ("Yes" in step S7), the control unit 8 records the reflection position and reflection intensity from the identified landmark 3 (step S9). Next, when the distribution of the recorded positions exceeds the code display angle, the control unit 8 reads the code from the position of the recorded high reflection intensity area and returns to the road sign detection state (steps S11 to S13).

[0075] Specifically, it is determined whether "maximum angle of recorded points > reading angle" (step S10), and if "No" in step S10, the process returns to step S1. If "Yes" in step S10, the code 13 is read (step S11), and the own position is corrected (the own position is identified) based on the read code (step S12). Then, "code reading FLAG" is set to "FALSE" (step S13), and the process returns to step S1.

[0076] With reference to Figures 25 to 29, the maximum angle of a recorded point and the recording at the time of code reading will be described. Here, the scene shown in Figure 25 is assumed. Figure 25 illustrates the positional relationship between a landmark 3 (here, a road sign) and a detection unit 5 (here, Lidar). As shown in Figure 25, a scene is assumed in which Lidar passes in a straight line "5 m" away from the road sign, and the point closest to the road sign is set to point P0, which is "0", and it moves from point P1, which is "-15000 mm", to point P6, which is "+15000 mm".

[0077] For ease of explanation, let us assume that a landmark 3 shown in Fig. 26 is to be detected. The landmark 3 shown in Fig. 26 is a road sign, and its entire periphery is formed of a high reflection intensity area (one example is a retroreflective material). As shown in Fig. 26, the point at which the lidar approaches closest is designated as "0°," and the side that is visible as the lidar approaches is represented by a negative angle, and the side that is visible as the lidar moves away is represented by a positive angle.

[0078] Figure 27 is a graph showing the positions of recorded points at point P1 at "-15000 mm." The graph shown in Figure 27 was created based on the reflection intensity irradiated onto the range indicated by symbol Q1 in Figure 26. As shown in Figure 27, at point P1 at "-15000 mm," the maximum angle of recorded points is "60°."

[0079] FIG. 28 is a graph showing the positions of recorded points when the laser beam has moved to point P3 at "-5000 mm." The graph at point P2 at "-10000 mm" shown in FIG. 28 was created based on the reflection intensity irradiated in the range indicated by reference symbol Q2 in FIG. 26. The graph at point P3 at "-5000 mm" shown in FIG. 28 was created based on the reflection intensity irradiated in the range indicated by reference symbol Q3 in FIG. 26. As shown in FIG. 28, the position of the smallest recorded point at point P3 at "-5000 mm" is "-115°," but the value at the position of "-120°" was recorded at point P2 at "-10000 mm" and can be referenced. Therefore, the minimum recorded point value at point P3 of "-5000 mm" is "-120°" at point P2 of "-10000 mm", and the maximum value is "20°" at point P3 of "-5000 mm", so the maximum recorded point angle at point P3 of "-5000 mm" is "140°".

[0080] FIG. 29 is a graph of the positions of recorded points when the vehicle has moved to point P6 at "15,000 mm." FIG. 29 accumulates recorded points at point P1 at "-15,000 mm," point P2 at "-10,000 mm," point P3 at "-5,000 mm," point P0 at "0 mm," point P4 at "5,000 mm," point P5 at "10,000 mm," and point P6 at "15,000 mm." Note that, since the landmark 3 (see FIG. 26) is assumed to be entirely made of a high-reflection-intensity region (e.g., a retroreflective material), there are no recorded points with low reflection intensity in FIG. 29. However, if the landmark 3 has a low-reflection-intensity region, there will be recorded points with low reflection intensity.

[0081] 30 to 33, a case where the landmark 3 has a low reflection intensity region will be described. Fig. 30 shows an example of a landmark 3, where (a) is a code section 20 that represents the number "5" with a surface division number of "2", (b) is a code section 20 that represents the number "2" with a surface division number of "2", and (b) is a code section 20 that represents the number "6" with a surface division number of "2".

[0082] Fig. 31 shows the results of detection on the code section 20 representing the number "5" shown in Fig. 30(a). Fig. 32 shows the results of detection on the code section 20 representing the number "2" shown in Fig. 30(b). Fig. 33 shows the results of detection on the code section 20 representing the number "6" shown in Fig. 30(c). The results shown in Figs. 31 to 33 are the contents accumulated when moving from point P1 at "-15000 mm" to point P6 at "15000 mm".

[0083] If reading becomes impossible due to an increase in distance during recording, for example, it is advisable to terminate the reading state and return to the road sign detection state.The event of distance increase during recording is determined based on distance information obtained by Lidar or on-board sensors such as wheel speed and gyro sensors.

[0084] As described above, the landmark 3 according to this embodiment can be given, as a code pattern, the identification information and position information of the landmark 3. Therefore, by reading this code pattern, the landmark 3 can accurately recognize its own position. Furthermore, in the position estimation method using the landmark 3 according to this embodiment, it is possible to read the code pattern of the landmark 3 and obtain the identification information and position information of the landmark 3 from the read code pattern. Therefore, it is possible to recognize one's own position with high accuracy.

[0085] Although the embodiment of the present invention has been described above, the present invention is not limited to this and can be practiced within the scope of the claims.

[0086] In the embodiment, as shown in FIG. 1, outdoor position detection and autonomous movement are assumed. However, the autonomous mobile system 1 (including the position estimation system 1A) can also be applied indoors, such as in tunnels and warehouses. FIG. 34 shows an image of the autonomous mobile system 1 being used in a tunnel. FIG. 35 shows an image of the autonomous mobile system 1 being used in a warehouse.

[0087] In the embodiment, the case where the number of display digits of the code unit 20 is "3" has been described as an example (see FIG. 6). However, the number of display digits may be other than "3", and as shown in FIG. 36, for example, the number of display digits may be "2", "4", etc. FIG. 36 is a diagram for explaining variations in the number of display digits of the code unit 20.

[0088] In the embodiment, as shown in Fig. 5, the code section 20 is composed of a base code section 21, a second code section 22, and a third code section 23, so that the number of code sections is "3," and end emphasis sections 31 and 32 are provided on both sides of the code section 20. However, as shown in Fig. 37(a), it is also possible to configure the code section 20 without providing end emphasis sections 31 and 32 on both sides. Also, as shown in Fig. 37(b), it is also possible to set the number of code sections to a number other than "3" ("4" in Fig. 37(b)).

[0089] In the embodiment, the symbol portion 20 is attached to a cylindrical (including cylindrical, whether hollow or not) three-dimensional shaped portion. However, the cross-sectional shape of the three-dimensional shaped portion is not limited to a circle, and the cross-sectional shape of the three-dimensional shaped portion may be other than a circle (for example, an ellipse or a polygon).

[0090] In the embodiment, the landmark 3 is assumed to be a road signpost, and the code portion 20 is attached to the three-dimensional shape portion of the road signpost. However, the object to which the code portion 20 is attached is not limited to a road signpost, and various means can be used to attach the code portion 20. Furthermore, if the landmark 3 itself has a three-dimensional shape portion, the code portion 20 can be attached to the entire landmark 3.

[0091] A landmark 103 according to a first modified example will be described with reference to FIG. 38. FIG. 38 is a schematic diagram of the landmark 103 according to the first modified example, where (a) shows the landmark 103 in a disassembled state and (b) shows the landmark 103 in a state in use. As shown in FIG. 38, the landmark 103 includes a triangular cone 110, a pole 130, and a mounting fixture 140. The triangular cone 110 is an example of a cone-shaped three-dimensional portion (whether or not the interior is hollow). A hole 111 is formed at the apex of the triangular cone 110, and the pole 130 can be attached to the hole 111. The pole 130 is fixed to the triangular cone 110 via the mounting fixture 140. The mounting fixture 140 has a shape corresponding to the hole 111. A code portion 120 is formed on the periphery of the pole 130. The landmark 103 is easy to carry and can be used as an instant marker. The cross-sectional shape of the cone-shaped three-dimensional portion may be other than circular (e.g., elliptical or polygonal).

[0092] A landmark 203 according to the second modified example will be described with reference to Fig. 39. Fig. 39 is a schematic configuration diagram of the landmark 203 according to the second modified example. The landmark 203 shown in Fig. 39 is a triangular cone 210 itself to which a sign portion 220 is attached, and the sign portion 220 is formed on the cone-shaped portion.

[0093] A landmark 303 according to a third modified example will be described with reference to FIG. 40. FIG. 40 is a schematic diagram of the landmark 303 according to the third modified example, where (a) shows the landmark 303 in an exploded state and (b) shows the landmark 303 in a state in use. As shown in FIG. 40, the landmark 303 includes a triangular cone 310 and a cover 330. The cover 330 has a weight portion 340 and a covering portion 350. The covering portion 350 corresponds to the shape of the conical portion of the triangular cone 310 and can cover part or all of the conical portion. A code portion 320 is formed on the covering portion 350.

[0094] A landmark 403 according to the fourth modified example will be described with reference to FIG. 41. FIG. 41 is a schematic diagram of the landmark 403 according to the fourth modified example, where (a) shows the landmark 403 in a disassembled state and (b) shows the landmark 403 in a state in use. The landmark 403 shown in FIG. 41 has a configuration in which the weight portion 340 is removed from the landmark 303 according to the third modified example. The landmark 403 includes a triangular cone 410 and a cover 430. The cover 430 corresponds to the shape of the conical portion of the triangular cone 410 and can cover part or all of the conical portion. A code portion 420 is formed on the cover 430. [Explanation of symbols]

[0095] 1 Autonomous Mobile System 1A Positioning System 2. Mobile 3,103,203,303,403 Landmarks 4. Moving Part 5. Detection unit 6. Communications Department 7 Memory section 8 Control Unit 11 Support section 12 Extension 20 Sign part 21 Standard code part 22 2nd code part 23 3rd code part 30,31,32 Edge emphasis part

Claims

1. A landmark installed on a route traveled by a moving object, a columnar or conical three-dimensional shape portion; a code portion is provided on a surface of the three-dimensional shape portion, the code portion has a plurality of divided regions in a circumferential direction of the three-dimensional shape portion, and a code pattern is expressed by changing the reflection intensity for each divided region. A distinctive landmark.

2. the divided regions are either high reflection intensity regions or low reflection intensity regions, a binary code pattern is expressed by an arrangement pattern using the high reflection intensity regions and the low reflection intensity regions; The landmark of claim 1 .

3. The code portion has either a divided area to which one or both of a character and a symbol are added, or a divided area whose width is changed. The landmark of claim 1 .

4. the code unit further includes a reference code unit that specifies a reading range of the code pattern; The landmark of claim 1 .

5. The three-dimensional shape portion further has an end emphasis portion provided on one side or both sides of the code portion in the axial direction. The landmark of claim 1 .

6. 2. The method for estimating a position of a moving object using landmarks according to claim 1, a point cloud data acquisition step of acquiring point cloud data using a detection unit included in the moving object; a landmark detection step of detecting a point cloud corresponding to the landmark from the point cloud data; a code pattern reading step of reading the code pattern from the point cloud recognized as the landmark; a position specifying step of specifying a position based on the read code pattern. A position estimation method comprising:

7. In the code pattern reading step, when a reading angle becomes equal to or greater than a code display angle indicating an angle occupied by one of the code patterns, the code pattern is read.

7. The method of claim 6, wherein the location is estimated based on the location of the target object.

8. A program for causing a computer to execute the position estimation method according to claim 6.

Images