Method, system and sensor subsystem for detecting obstacles

Abstract

The application discloses a method and system for detecting obstacles for autonomous navigation machines. The method is executed in a system, and the system has a light emitter, a light sensor, an image processor and a central processing unit. The light emitter and the light sensor are arranged on the device at a distance. When the light emitter emits an indicator light and projects it on the path of the autonomous navigation machine, the light sensor senses the indicator light and generates an image including the indicator light. After analyzing the image, at least one feature of the indicator light can be obtained. This feature can be used to obtain a spatial relationship between the autonomous navigation machine and the obstacle. The spatial relationship will allow the system to determine whether the autonomous navigation machine will collide with a wall or fall off a cliff.

Description

[0001] This application is a divisional application of the invention patent application "Method, system and sensor subsystem for detecting obstacles" with application date of May 23, 2018 and application number 201810503115.0. Technical Field

[0002] The application discloses a technology for detecting obstacles, particularly a method for detecting obstacles such as cliffs on the path of a mobile machine, a system for implementing the method, and a related sensor subsystem. Background Art

[0003] Autonomous navigation devices that can autonomously determine their travel paths, such as automatic vehicles or automatic sweeping robots, need to avoid collisions with other objects or the risk of falling from a height while traveling. For example, in order to avoid colliding with a wall, the autonomous navigation device needs a proximity sensor to determine the distance to the wall so as to avoid collision. Existing proximity sensors are radars, ultrasonic sensors, or optical sensors, such as infrared sensors, which can determine nearby objects without touching them.

[0004] For example, an ultrasonic sensor emits ultrasonic waves, and determines the object in front by receiving the reflected sound waves from the object. The distance can be obtained by the time difference between the time of emitting the ultrasonic wave and the time of receiving the reflected sound wave. For cliff detection, a cliff sensor installed under an autonomous navigation device generally uses infrared light or ultrasonic waves, and can sense the ground below the device by measuring the reflected or scattered light or sound waves. Therefore, the cliff sensor can prevent the autonomous navigation device from moving to the cliff and falling.

[0005] When the autonomous navigation device can detect obstacles such as the wall or cliff mentioned above, the controller of the device can obtain results from the sensing data and instruct the drive system in the device to stop or avoid the obstacle when the autonomous navigation device approaches or reaches the obstacle. Summary of the invention

[0006] One of the purposes of the obstacle detection method and system disclosed in the application is to detect obstacles on the path of an autonomous navigation machine. When the autonomous navigation machine is running, the obstacle detection method is executed by the system of the autonomous navigation machine itself, wherein an algorithm is used to process data generated by the sensor subsystem in the device.

[0007] According to one embodiment of the sensor subsystem, a system for performing the method of detecting an obstacle includes a light emitter and a light sensor, wherein the light emitter emits an indicator light, and the light sensor captures an image including the indicator light at a time, or captures a sequence of images generated over a period of time. In one embodiment, information captured from the image can be used to estimate the distance to the obstacle. In one method, the change of the indicator light image can be obtained from a sequence of images, and it can be used to estimate whether the autonomous navigation machine is approaching the obstacle, so that the autonomous navigation machine can process a collision avoidance measure to avoid the risk of collision.

[0008] The indicator light emitted from the light transmitter is like a linear light projected on the area in front of the autonomous navigation machine, so the light sensor on the device can capture an image including this linear light. After the image analysis program, information for judging the distance to the obstacle can be generated from the linear light image. The information can reflect how the autonomous navigation machine approaches the obstacle.

[0009] The obstacle detection method is applied to an autonomous navigation machine equipped with a light emitter and a light sensor. The light emitter and the light sensor are arranged on the autonomous navigation machine, preferably at a distance from each other. The light emitter emits an indicator light projected onto the path of the autonomous navigation machine, and the light sensor is used to sense the indicator light and generate an image with the indicator light.

[0010] After analyzing the image, at least one feature of the sensed indicator light can be obtained, so that the system can obtain the spatial relationship between the autonomous navigation machine and the obstacle based on the feature of the indicator light. This spatial relationship allows the autonomous navigation machine to calculate the distance to the obstacle and determine whether the autonomous navigation machine will collide with the obstacle based on the comparison result with a collision threshold, or determine whether the autonomous navigation machine will fall into the obstacle, that is, the cliff, based on the comparison result with a fall threshold.

[0011] According to another embodiment, a system for detecting obstacles installed in an autonomous navigation machine includes a controller, a light emitter, a light sensor, an image processor and a central processing unit, wherein the image processor is used to generate an image including indicator light projected onto the path of travel of the autonomous navigation machine, and the central processing unit is used to execute a computer program related to the obstacle detection method.

[0012] The system has a sensor subsystem. In one embodiment, the sensor subsystem mainly includes a light emitter, a light sensor and an image processor, which performs projection of an indicator light on the path of the autonomous navigation machine, senses the indicator light, and generates an image including the indicator light. The indicator light is provided to the autonomous navigation machine for analysis to obtain the spatial relationship between the autonomous navigation machine and the obstacle.

[0013] Furthermore, after obtaining the spatial relationship, the system can calculate the distance between the autonomous navigation machine and the obstacle, and determine whether the autonomous navigation machine will collide with the obstacle after comparing it with a collision threshold recorded in the system memory, or determine whether the autonomous navigation machine will fall into the obstacle after comparing it with a drop threshold recorded in the memory. When the system determines that the autonomous navigation machine will reach the collision threshold or drop threshold, the central processor of the system will generate a signal to instruct the controller to drive the autonomous navigation machine to avoid the obstacle. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A schematic top view of an embodiment of a scenario showing an autonomous navigation machine approaching an obstacle;

[0015] Figure 2 A schematic diagram of an embodiment of the situation showing an autonomous navigation machine in front of a cliff;

[0016] Figure 3 A perspective schematic diagram showing an embodiment of an autonomous navigation machine in front of a wall;

[0017] Figure 4A and Figure 4B A schematic diagram showing an embodiment of an autonomous navigation machine approaching a wall;

[0018] Figure 5A and Figure 5B A schematic diagram showing an embodiment of an autonomous navigation machine approaching a cliff;

[0019] FIG. 6A to FIG. 6D A schematic diagram showing an embodiment of an autonomous navigation machine approaching an obstacle at a height above the ground;

[0020] Figure 7 A circuit block diagram showing an embodiment of an obstacle detection system;

[0021] Figure 8 A schematic diagram of an embodiment of using a light sensor to obtain an indication of light changes when an autonomous navigation machine approaches a wall is shown;

[0022] Fig. 9 A schematic diagram showing an embodiment of using a light sensor to obtain an indication of light changes when an autonomous navigation machine approaches a cliff;

[0023] FIG. 10A to FIG. 10C A schematic diagram showing an embodiment of an autonomous navigation machine approaching an obstacle and sensing light;

[0024] Fig.11 A schematic diagram showing an embodiment of a circular indicator light projected on a wall when an autonomous navigation machine approaches the wall;

[0025] Fig.12A flowchart of an embodiment of an obstacle detection method for an autonomous navigation machine is shown;

[0026] Fig.13 A flow chart showing yet another embodiment of an obstacle detection method applied to an autonomous navigation machine. DETAILED DESCRIPTION

[0027] The invention disclosed in the present disclosure is about an obstacle detection method and an obstacle detection system applied to an autonomous navigation machine, which is like an unmanned vehicle or a cleaning robot that can automatically cruise in an area. The obstacle detection method running in the autonomous navigation machine allows the autonomous navigation machine to sense obstacles on the travel path and identify obstacles in front of the machine, so as to automatically drive the autonomous navigation machine to cruise. The obstacle detection system is like a sensor subsystem running in the autonomous navigation machine, which mainly includes a light emitter and a light sensor installed in the machine.

[0028] The sensor subsystem is characterized in that the light emitter and the light sensor are devices that are horizontally arranged and separated by a distance. According to the embodiment, the light emitter and the light sensor can be arranged at the same or different horizontal positions on a plane. The light emitter can use a light source such as a laser or a light emitting diode (LED) that can emit linear light, and the emitted linear light is used as the indicator light of the autonomous navigation machine. The indicator light can also be a light emitting with a specific area. The light sensor is used to sense the linear light emitted by the light emitter on the path of the autonomous navigation machine, or the light projected with a certain area.

[0029] According to one of the purposes of the obstacle detection method and system disclosed in the present invention, the method is used to detect obstacles on the path of an autonomous navigation machine, such as a wall, a cliff, or a floating obstacle at a height above the ground. The method is implemented by an algorithm operating in an obstacle detection system, so that the autonomous navigation machine can process the data generated by the above-mentioned sensor subsystem by itself to avoid obstacles on the path ahead of the autonomous navigation machine.

[0030] according to Figure 1 In the illustrated embodiment, this schematic diagram is a top view illustrating an autonomous navigation machine approaching an obstacle while traveling.

[0031] The autonomous navigation machine 10 in the figure can be an autonomous robot, such as a cleaning robot, etc., which can cruise in areas with various terrains. The obstacle detection system installed in the autonomous navigation machine 10 is used to sense obstacles on the path of the autonomous navigation machine 10. The system is such as a sensor subsystem installed in the autonomous navigation machine 10, which includes the light emitter 101 and light sensor 102 shown in the figure, and related processing circuits. The light emitter 101 and light sensor 102 in this example are devices separated by a distance on the horizontal plane. However, the light emitter 101 and light sensor 102 do not necessarily have to be located at the same horizontal height.

[0032] The light emitter 101 can be a laser or a light emitting diode as a light source to emit a linear light 103. The linear light 103 can be formed by the light emitted by the light source passing through a specific lens. When the autonomous navigation machine 10 moves, the linear light 103 will continue to be projected on the path of the autonomous navigation machine 10, such as the ground 19 in the figure or a surface with a specific terrain, generally projected in front of the autonomous navigation machine 10. For example, when the autonomous navigation machine 10 runs close to the wall 18, the wall 18 forms an obstacle at a distance in front of the autonomous navigation machine 10, and the linear light 103 projected by the autonomous navigation machine 10 will form an indicator light projected on the wall 18 and the ground 19. According to the content of the figure, a boundary 15 is formed between the ground 19 and the wall 18, which divides the linear light 103 into a first line segment (such as the line segment at the bottom of the figure) and a second line segment (such as the line segment at the top of the figure), and these two line segments will be sensed by the light sensor 102 on the autonomous navigation machine 10. The optical sensor 102 is like a camera installed on the autonomous navigation machine 10, and has a viewing angle, such as the range indicated by two dotted lines in the figure, and the optical sensor 102 is used to sense (photograph) the image in this area. It is mentioned here that the viewing angle of the optical sensor 102 should cover the indicator light 13 to effectively detect the object in front of the autonomous navigation machine 10. In this example, the optical sensor 102 is used to sense the linear light 103 and photograph the image including the linear light 103. When the obstacle detection system in the autonomous navigation machine 10 is running, the lower line segment of the linear light 103 can be sensed by the optical sensor 102. For example, the length of the lower line segment of the linear light 103 can be obtained by the optical sensor 102, and this length change is a reference for determining the distance "d" to the obstacle.

[0033] It is worth mentioning that, in addition to the length of the light obtained from the optical sensor 102, the horizontal position of the upper line segment of the linear light 103 can also be used as another basis for determining the distance "d". In one embodiment, the optical sensor 102 can continuously sense the indicator light projected by the light emitter 101, so that the sensor subsystem generates a sequence of images including the indicator light. If any changes are detected in the image length on the line segment of the lower linear light 103, or any changes are detected in the position or length of the upper line segment from these images, the system can know whether the autonomous navigation machine 10 is approaching or moving away from the wall 18 according to the changes.

[0034] In addition to the above-mentioned length or position change of the indicator light image obtained by the light sensor 102, any length or position of the indicator light that can be obtained from the continuous image can be used as a basis for judging the state of the autonomous navigation machine 10. For example, the movement trend of the autonomous navigation machine 10 can be obtained through the changes in the length and position of the light in the continuous image. The obstacle detection system running therein establishes an alarm mechanism based on this, and can issue an early warning based on the spatial relationship between the autonomous navigation machine 10 and its obstacle (such as the wall 18).

[0035] Another form of the obstacle is a cliff, which can be seen in Figure 2 The illustrated schematic diagram depicts an embodiment of a situation in which an autonomous navigating machine is in front of a cliff.

[0036] In this example, the autonomous navigation machine 10 cruises on a plane 20 with some terrain, such as running on a table or the ground, and there is a cliff 22 in front, which can be formed by the vertical part of the table edge or the stairs on the ground. The light emitter 101 of the autonomous navigation machine 10 continuously emits the indicator light 103' projected on the path of the autonomous navigation machine 10. The indicator light 103' shown in this example is shown as a vertical linear light emitted by the light emitter 101. In the figure, a certain line segment represents a certain range of the emitted light. When the light sensor 102 senses the indicator light 103' within its viewing angle, the system will generate an image including the indicator light 103'.

[0037] In one embodiment, the indicator light 103' is shown to be divided into two by the edge 15' between the plane 20 and the cliff 22. The indicator light 103' in the image sensed by the optical sensor 102 shows that the autonomous navigation machine 10 will encounter an obstacle, that is, the cliff 22 in this example. The length, slope or angle of the line segment in the image sensed by the optical sensor 102 can allow the system to determine the distance between the autonomous navigation machine 10 and the obstacle represented by the edge 15'. Therefore, when the obstacle detection system identifies the cliff 22 on the path of the autonomous navigation machine 10, the obstacle detection system running in the autonomous navigation machine 10 will instruct the machine to avoid the cliff 22.

[0038] Figure 3 Another schematic diagram of an embodiment is shown in which the autonomous navigation machine 10 has a wall as an obstacle in front.

[0039] The figure shows that the autonomous navigation machine 10 is cruising on the ground 30 and moving towards a wall 32. During this process, the light emitter 101 emits linear light as an indicator light in the vertical direction. The emitted light will be projected onto the ground 30 and the wall 32 in front of the autonomous navigation machine 10 at the same time.

[0040] When the autonomous navigation machine 10 approaches the wall 32, the line segment of the indicator light 103" indicates the degree of proximity of the autonomous navigation machine 10 to the wall 32. In the obstacle detection method, the slope and angle of the indicator light 103" can indicate the spatial relationship between the autonomous navigation machine 10 and the wall 32. Alternatively, the position of the line segment of the indicator light projected on the wall 32 can also be used as an indicator to describe the spatial relationship.

[0041] According to the above embodiments, the light emitter emitting linear light in the obstacle detection system is installed on the autonomous navigation machine, and the linear light forms an indicator light. The characteristics of the indicator light become the basis for judging the spatial relationship between the autonomous navigation machine and the obstacle (wall, cliff, etc.). Therefore, the obstacle detection system can effectively prevent the autonomous navigation machine from colliding with the wall or falling off the cliff. Among them, because the light emitter and the light sensor on the autonomous navigation machine are designed to be horizontally (but not limited to) separated by a certain distance, when the indicator light is sensed by the light sensor, the light characteristics obtained can be analyzed to generate information such as length, position, slope / angle, and / or area, and at least one of the characteristics is sufficient to judge the spatial relationship between the autonomous navigation machine and the obstacle. The changes in these characteristics obtained from multiple images with indicator light are sufficient to generate the movement trend of the autonomous navigation machine, so that the obstacle detection system can issue a collision warning or a fall warning in time.

[0042] Figure 4A and Figure 4B A schematic diagram showing an embodiment of any characteristic change that causes an obstacle detection system in an autonomous navigation machine to issue a wall collision warning.

[0043] exist Figure 4AIn the side view, an autonomous navigation machine 40 (such as an automatic vehicle) is shown cruising on a ground 43. The light emitter of the obstacle detection system in the autonomous navigation machine 40 emits an indication light projected in front of the machine. The light emitted by the light emitter has a certain light range 42, and the emitted light is limited to this range. When the light sensor in the obstacle sensing system senses the indication light, the system can determine the spatial relationship between the autonomous navigation machine 40 and the obstacle (wall 44) at a certain distance. When the indication light is projected on the ground 43 and the wall 44 at the same time, the indication light sensed by the light sensor in its field of view is divided into a first line segment with a length of "40a" and a second line segment with a length of "40b". The sensed image also includes these two line segments. Then, the obstacle detection system calculates and analyzes the image including all or part of the indication light.

[0044] When the autonomous navigation machine 40 approaches the wall 44, Figure 4B , showing a gradually smaller light range 42' relative to the light range 42, and at the same time, this change is sensed by the optical sensor of the system, and the first line segment of the indicator light projected on the ground 43 changes from "40a" to "40a'", and the second line segment of the indicator light projected on the wall 44 changes from the height "40b" to "40b'". In this way, the obstacle detection system uses the change characteristics of the first line segment or the second line segment to determine the spatial relationship between the autonomous navigation machine 40 and the obstacle (wall 44), such as the distance in this example.

[0045] Furthermore, when the system captures two or more images including indicator light, any characteristic change of the indicator light can be used to obtain the movement trend of the autonomous navigation machine 40, and this characteristic can be used to determine whether it is approaching an obstacle.

[0046] Figure 5A and Figure 5B A schematic diagram of an embodiment schematically showing any characteristic change that enables a system of an autonomous navigation machine to issue a cliff fall warning.

[0047] like Figure 5A The schematic side view of the autonomous navigation machine 50 cruising on the ground 53 is shown. In this example, there is a cliff 54 at a distance from the autonomous navigation machine 50. The light emitter of the obstacle detection system emits an indication light, which is transmitted on the ground 53 in front of the autonomous navigation machine 50, and then the indication light is sensed by the optical sensor of the system. When the autonomous navigation machine 50 is a certain distance away from the cliff 54, the indication light includes a line segment projected on the ground 53 and an indication light segmented by the edge between the ground 53 and the cliff 54. When the indication light is projected on the ground 53, the light at a distance "50a" is sensed by the optical sensor. This image including all or part of the indication light will be calculated and analyzed by the system.

[0048] In this example, the autonomous navigation machine 50 approaches a cliff 54. Figure 5B As shown, the length of the indicator light projected on the ground 53 becomes shorter, and the length of the indicator light sensed by the light sensor on the ground 53 is Figure 5A The indicator light will be “50a” and then “50a′”. Therefore, the obstacle detection system can determine the distance between the autonomous navigation machine 50 and the obstacle (cliff 54) according to the length characteristics of the indicator light.

[0049] Similarly, when the system captures two or more images including the indicator light, the change in the length of the indicator light can be used to obtain the movement trend of the autonomous navigation machine, and then determine whether the autonomous navigation machine is approaching an obstacle.

[0050] FIG. 6A to FIG. 6D A schematic diagram showing an embodiment of an autonomous navigation machine approaching a floating obstacle of a certain height.

[0051] exist Fig. 6A In the example, the autonomous navigation machine 60 cruises on the ground 63 and approaches a floating obstacle 64 at a certain height from the ground 63. The optical sensor of the autonomous navigation machine 60 senses the indicator light projected onto the floating obstacle 64. The indicator light is a linear light, which is emitted by the optical transmitter in this example. When the autonomous navigation machine 60 approaches the floating obstacle 64 at a certain height from the ground 63 in front, the indicator light forms two line segments on the ground 63 and the floating obstacle 64, wherein the indicator light projected onto the floating obstacle 64 forms an image of length "60b" in the optical sensor.

[0052] When the obstacle detection system is running, the optical sensor continuously captures the indicator light in front of the autonomous navigation machine 60, including the indicator light line segment formed on the ground 63 and another line segment of length "60b" formed on the floating obstacle 64. Figure 6B Specifically shown is the image marked in the frame 66. The frame 66 shows at least two image features, wherein the first line segment 60a represents the line segment of the indicator light projected on the ground 63, and the second line segment 60b in the frame 66 represents the line segment of the indicator light projected on the floating obstacle 64.

[0053] Then, if Figure 6C The autonomous navigation machine 60 gradually approaches the floating obstacle 64 above the ground 63. The floating obstacle 64 causes the range of light 62' emitted by the light transmitter to change, that is, the two line segments projected by the indicating light to the ground 63 and the floating obstacle 64 change. For example, because of the spatial relationship between the autonomous navigation machine 60 and the floating obstacle 64, the line segment projected to the floating obstacle 64 changes from "60b" to "60b'".

[0054] The image captured by the optical sensor is as follows Fig.6D The frame 66' in the figure shows that when the autonomous navigation machine 60 approaches the floating obstacle 64, the first line segment 60a' and the second line segment 60b' will change. For example, the figure shows that the length and slope of the first line segment 60a' have changed. The length of the first line segment 60a' is shorter than that of the first line segment 60a. When the autonomous navigation machine 60 approaches the floating obstacle 64, the slopes of the first line segments 60a, 60a' also change. Moreover, the second line segments 60b, 60b' also show changes in length and position. For example, the length of the second line segment 60b' is shorter than that of the second line segment 60b, and the position also moves to the left.

[0055] According to another embodiment, the sensor module in the obstacle detection system mainly includes a light emitter and a light sensor arranged at a certain distance. The light emitter emits a specific indicator light, and the light sensor captures an image including the indicator light once or continuously for a period of time. The indicator light is such as a linear light, a circular light, or any shape of light projected in front of the autonomous navigation machine. The light sensor is used to obtain images including these indicator lights. Through image analysis, the characteristics of these indicator lights can be obtained to determine the spatial relationship between the autonomous navigation machine and the obstacle. This spatial relationship can allow the obstacle detection system in the machine to indicate and drive the autonomous navigation machine to avoid the risk of collision and falling.

[0056] Figure 7 A circuit block diagram showing an embodiment of an obstacle detection system.

[0057] The obstacle detection system is equipped with a sensor subsystem, a processor and memory, and has data processing capabilities. The sensor subsystem is used to obtain environmental information around the autonomous navigation machine and establish a protection mechanism after processing the environmental information. The environmental information is the image generated by the obstacle detection system installed in the autonomous navigation machine, which can be used to determine whether the autonomous navigation machine will encounter the possibility of damage such as collision or falling.

[0058] As shown in the figure, the circuit block of the obstacle detection system includes a controller 701 for controlling the operation of each circuit component in the system. The controller 701 is used to drive the light emitter 702 to emit the indicator light and control the light sensor 705 to sense the light signal within its visual angle. The indicator light can be, for example, linear light, circular light or light of any shape. According to the control signal, the light can be emitted all the time or periodically. The light emitter 702 included in the system is coupled to the controller 701, and has a light source and a driving circuit. The light is emitted through necessary optical components and / or windows provided on the autonomous navigation machine. The optical components, such as lenses, can guide the indicator light to be linear light, circular light or light of any shape. The system's light sensor 705 is coupled to the controller 701. The light sensor 705 is at a distance from the light emitter 702 and is used to sense the light emitted by the light emitter 702 to obtain images on the path of the autonomous navigation machine. The image signal is transmitted to the image processor 704 to generate images including indicator light. The image processor 704 is coupled to the light sensor 705 to generate sensing images. These image signals are first temporarily stored in a memory, such as a memory 706.

[0059] According to an embodiment of the circuit planning, the main circuits in the autonomous navigation machine are a light emitter 702, a light sensor 705 and an image processor 704 in a sensor subsystem. The sensor subsystem is used to generate an indicator light and generate an image including the indicator light through image sensing. The autonomous navigation machine can use at least one feature on the indicator light image to obtain the spatial relationship between the autonomous navigation machine and an obstacle approaching the path of travel.

[0060] After at least one feature in the indicator light image is processed by the central processor 703 coupled to the image processor 704 and the controller 701, the central processor 703 is used to execute the obstacle detection method. In one embodiment, the memory 706 coupled to the central processor 703 operates as a system memory or a storage for storing instructions for executing the system function, for example, an instruction set for executing the obstacle detection method. The method executed by the central processor 703 mainly includes analyzing the image with the indicator light obtained by the light sensor 705, obtaining at least one feature in the indicator light image, and obtaining the spatial relationship between the autonomous navigation machine and the obstacle according to the feature in the indicator light image.

[0061] According to one embodiment, the features obtained from the image by the sensor subsystem can be used to estimate the distance between the autonomous navigation machine and the obstacle, that is, the spatial relationship between the two, so that the obstacle detection system can determine whether the autonomous navigation machine is in a dangerous situation, such as colliding with a wall or falling off a cliff. In one example, the controller 701 in the system is connected to the machine driver 708 of the autonomous navigation machine, and the machine driver 708 is connected to the drive system of the system. When the system determines that there is an obstacle in front of the autonomous navigation machine, the controller 701 generates a signal indicating that the machine driver 708 can respond to the obstacle.

[0062] In one embodiment, the light sensor 705 can continuously obtain a sequence of images with indicator light and sense changes in the continuous images to determine that the autonomous navigation machine is approaching an obstacle, so that the autonomous navigation machine can take measures to avoid the danger.

[0063] The following figures provide schematic diagrams of embodiments of the obstacle detection method disclosed in the present disclosure.

[0064] For example, when an autonomous navigation machine approaches a wall, Figure 8 As shown, a schematic diagram of an embodiment of the change of the indicator light image sensed by the light sensor is described.

[0065] The obstacle detection system installed in the autonomous navigation machine is a sensor subsystem as described in the above embodiment, wherein the main circuit is the above-mentioned light emitters 801, 801' and light sensors 802, 802'. The autonomous navigation machines 80, 80' shown operate like autonomous vehicles, and include a calculator, which drives the light emitters 801, 801' to emit indicator light after calculation, and integrates the data generated by the light sensors 802, 802' to process and generate data of terrain information on the path of the autonomous navigation machine 80.

[0066] As shown in the figure, the autonomous navigation machine 80 indicated by the solid line runs on the ground 83 and is first in the first position. The autonomous navigation machine 80 is provided with a light emitter 801 and a light sensor 802. The light emitter 801 emits an indication light 803. The indication light 803 shown in the figure is the imager 802 sensed by the light sensor 802 in its field of view. The indication light 803 shown has a turning point due to the boundary between the ground 83 and the wall 84, which divides the indication light 803 into two line segments.

[0067] The autonomous navigation machine 80 then moves from the first position to the second position and is closer to the wall 84, as shown by the autonomous navigation machine 80' in the dotted line in the figure. At this second position, the light emitter 801' still continues to emit the indicator light 803', and the light sensor 802' continues to sense the indicator light 803' and generate an image. Similarly, the indicator light 803' shown in the figure is the image sensed by the light sensor 802' in its perspective. The boundary between the ground 83 and the wall 84 causes the indicator light 803' to have a turning point and be divided into two line segments.

[0068] The example shows that when the autonomous navigation machine 80, 80' moves and approaches an obstacle (such as the wall 84 in this example), the indicator light emitted by the light emitter 801, 801' forms the indicator light 803, 803' line segments projected on the ground 83 and the wall 84, and the light sensors 802, 802' on the autonomous navigation machine 80, 80' sense the indicator light 803, 803', and in this example, linear light is formed projected on the ground 83 and the obstacle (such as the wall 84 in this example). The light sensor subsystem can obtain at least one feature of the indicator light 803, 803' through image processing analysis, for example, the length of the first line segment of the indicator light 803, 803' projected on the ground 83, and this example shows that when the autonomous navigation machine 80, 80' approaches the wall 84, the length of the first line segment becomes shorter when it reaches the second position. In addition, the slope of the first line segment of the indicator light 803, 803' can also be a feature for detecting obstacles. This example shows that when the autonomous navigation machine 80, 80' is closer to the wall 84, the slope of the first line segment becomes larger.

[0069] Furthermore, the second line segment of the indicator light 803, 803' projected on the wall 84 can also be another feature for determining the detected obstacle. For example, when the autonomous navigation machine 80, 80' is closer to the wall 84, the second line segment will become shorter. Thus, the lengths of the first line segment and the second line segment of the indicator light 803, 803' projected on the ground 83 and the wall 84 respectively can be used as the basis for detecting obstacles. In addition, the position of the second line segment and / or the slope or angle of the first line segment can also be used as the system's features for detecting obstacles.

[0070] Fig. 9 The example shown shows the change of the indicator light sensed by the light sensor when the autonomous navigation machine approaches the cliff. In this example, the autonomous navigation machine 90, 90' is running on the ground 94 and will approach an obstacle, which is a cliff (boundary 93) in this example, and the obstacle detection system in the autonomous navigation machine 90, 90' will be used to detect the obstacle and make a procedure to avoid falling.

[0071] The figure shows that the autonomous navigation machine 90, which is depicted by solid lines, is located at a first position, wherein the light emitter 901 emits an indication light (903, 904) at the first position, and then the light sensor 902 obtains an image of the indication light (903, 904) within a certain viewing angle. Next, the autonomous navigation machine 90 moves to a second position closer to the boundary 93, in this case, the autonomous navigation machine 90' ​​is depicted by dotted lines. The light sensor 902' of the autonomous navigation machine 90' ​​at the second position senses the indication light (903', 904') emitted by the light emitter 901', and the boundary 93 is formed by a cliff, and the indication light is divided into line segments projected on the ground 94, which are marked as first line segments 903, 903'. Due to the cliff, the other line segments 904, 904' into which the indication light is divided will not be connected to the first line segment (903, 903') projected on the ground 94, but will be projected on a wall some distance away from the cliff, and may still be sensed by the light sensors 902, 902'.

[0072] This example shows that the length changes of the first line segments 903 and 903' of the indicator light are sensed at different positions by the optical sensors 902 and 902', that is, because the autonomous navigation machine 90 and 90' moves from the first position to the second position, the first line segments 903 and 903' become shorter and have a greater slope when they are closer to the cliff boundary 93. Therefore, the characteristic changes of the first line segments 903 and 903' in the indicator light can be used as information for the system to detect obstacles, so that the obstacle detection system can obtain the result that the autonomous navigation machine 90 and 90' is approaching the cliff boundary 93.

[0073] Fig. 10A , 10B 10C schematically shows three image frames of the autonomous navigation machine approaching an obstacle at distances of 20 cm, 10 cm, and 5 cm.

[0074] Fig. 10A An image frame 1001 is displayed with a width of pixel 0 to pixel 200. The image frame 1001 displays a light segment 1004 indicating the light projected on the path of the autonomous navigation machine. The light segment 1004 displayed in this example represents the situation when the autonomous navigation machine is 20 centimeters away from an obstacle (a cliff in this example). The image frame 1001 also uses a dotted line 1005 to represent the light segment projected on the wall.

[0075] Fig. 10B Another image frame 1002 of the same size (width) as the image frame 1001 is displayed. The light segment 1006 displayed in the image frame 1002 indicates the light segment sensed by the light sensor when the light is projected on the path of the autonomous navigation machine at a distance of 10 cm from the obstacle. The dotted line 1007 represents the image of the light segment projected on the wall. The slope of the light segment 1006 is Fig. 10AThe light line segment 1004 in the middle image frame 1001 is larger, and the diagram shows that when the autonomous navigation machine is closer to the obstacle, the length of the line segment is shorter in the image frame of the same size.

[0076] Fig. 10C Another image frame 1003 of the same size is displayed, in which a light segment 1008 indicating that the light is projected on the path of the autonomous navigation machine at a distance of 5 cm from the obstacle is displayed, wherein a dotted line 1009 indicates that the light is projected on the wall. Similarly, the slope of the light segment 1008 is greater than Fig. 10B As shown in FIG. 10A , when the autonomous navigation machine is closer to the obstacle, the length of the segment 1008 also appears shorter than the segment 1006 in FIG. 10B in the image frame of the same size.

[0077] On the other hand, according to this example, Fig. 10A The dotted line 1005 is shown. Fig. 10B The dashed line 1007 and Fig. 10C As shown in the middle dotted line 1009, when the autonomous navigation machine approaches an obstacle (such as a wall), the light segment sensed by the light projected on the surface of the obstacle will gradually move to the left. Therefore, the position of the light segment can also be used as a basis for indicating the spatial relationship between the autonomous navigation machine and the obstacle.

[0078] The indicator light emitted by the light emitter can be a circular light, such as Fig.11 Schematic diagram of a circular indicator light projected onto a wall approached by an autonomous navigation machine.

[0079] The circular indicator light projected on the wall is sensed by a light sensor of the autonomous navigation machine that is spaced a certain distance from the light emitter. The sensed circular indicator light image is shown in the figure, and there will be some deformation. The circular indicator light image can specify a reference point (1111, 1112), and the projected image is divided into a first area (111, 113) below and a second area (112, 114) above by a dividing line, and the dividing line can be the boundary between the ground and the wall or cliff.

[0080] As shown in the figure, the solid circle represents the image sensed by the indicator light projected by the autonomous navigation machine at the first position, and the dotted circle represents the image generated by the autonomous navigation machine at the second position closer to the obstacle (such as a wall). This example shows that when the autonomous navigation machine approaches the obstacle, the first and second areas presented by the dotted circle are smaller than the solid circle. With reference to the reference points 1111 and 1112 corresponding to the solid and dotted circles, when the autonomous navigation machine approaches the obstacle, the dotted circle moves to the left.

[0081] According to the above implementation examples, the area and position of the circular indicator light can serve as an indicator basis for determining the spatial relationship between the autonomous navigation machine and the obstacle.

[0082] When the system thus obtains the spatial relationship between the autonomous navigation machine and the obstacle, the CPU in the obstacle detection system will calculate the distance between the autonomous navigation machine and the obstacle, and thus can determine whether the autonomous navigation machine will collide with the obstacle based on the result of the collision threshold stored in the system memory. Similarly, the spatial relationship allows the obstacle detection system to determine whether the autonomous navigation machine will fall under the obstacle based on the comparison with a drop threshold in the memory.

[0083] As mentioned above Figure 7 In the described system embodiment of the autonomous navigation machine, an obstacle detection method is performed, wherein the method steps are implemented such as Fig.12 Flowchart shown.

[0084] In step S121, the light emitter of the system emits an indicator light, and in step S123, the light sensor of the system is used to sense the indicator light and generate an image. The light sensor is like a camera, which is used to capture the indicator light within its field of view and form an image. The indicator light is projected on the path of the autonomous navigation machine, which may include the ground and / or the obstacle itself, so that the indicator light image can be used to reflect the spatial relationship between the autonomous navigation machine and the obstacle. The image processor in the system will be used to analyze at least one feature of the indicator light in the image, as shown in step S125.

[0085] According to the above embodiments, the characteristics of the indicator light, such as length, position, slope and / or area, any characteristic of the indicator light can be processed by the central processing unit to calculate the length, position and / or slope of a certain line segment of the sensed indicator light, or calculate the area of ​​the circular indicator light, such as step S129. One or a combination of these characteristics can allow the system to determine the distance between the autonomous navigation machine and the obstacle.

[0086] Therefore, the system can also obtain the movement trend of the autonomous navigation machine based on the characteristic changes of the indicator light obtained over a period of time. The implementation process is shown in Figure 13.

[0087] In step S151, the light emitter continuously emits the indicator light and projects it onto the ground and / or obstacles in front of the autonomous navigation machine. As in step S153, the light sensor on the autonomous navigation machine is driven to take at least two images including the indicator light in front of the autonomous navigation machine. After analyzing the images within a time range, as in step S155, the system can thus obtain at least one feature of each image, such as the length, position and / or slope, or the area of ​​the circular indicator light. As in step S157, the changes in the length, slope and / or area of ​​the indicator light in these images obtained over a period of time can be used to determine the changes in the spatial relationship between the autonomous navigation machine and the obstacles.

[0088] The changes in the indicator light projected on the ground and / or obstacles within a time range can be used as a basis for judging the movement trend of the autonomous navigation machine, so that the system can determine whether the autonomous navigation machine is gradually approaching an obstacle and respond, including generating warning information and avoiding it.

[0089] According to the above embodiments, the obstacle detection method and system are operated in an autonomous navigation machine such as an autonomous vehicle or a cleaning robot, wherein the system can derive the spatial relationship between the autonomous navigation machine and the obstacle based on the characteristics of the light emitted on the travel path. This spatial relationship allows the system in the autonomous navigation machine to calculate the distance to the obstacle, and thus determine whether the autonomous navigation machine will collide with the obstacle (such as a wall) or fall under the obstacle (such as a cliff). This provides the autonomous navigation machine with an alarm measure that can issue a warning in a timely manner when the autonomous navigation machine approaches an obstacle and drive the system to avoid it.

[0090] The above description is only a preferred embodiment of the present invention, which does not limit the patent scope of the present invention. Therefore, all equivalent structural changes made by using the description and illustrations of the present invention are also included in the scope of the present invention and are hereby stated.

Claims

1. An obstacle detection method, applied to an autonomous navigation machine, the autonomous navigation machine comprising a light transmitter and a light sensor separated by a distance, characterized in that: The method includes: The light emitter emits an indication light projected onto a travel path of the autonomous navigation machine, wherein the indication light is a vertical linear light; The light sensor senses the indicator light projected on the travel path to generate an image including the indicator light, wherein the light sensor is used to sense the vertical linear light, and at least one characteristic of the vertical linear light is the length, position or slope of the vertical linear light; In the autonomous navigation machine, analyzing the image to obtain at least one characteristic of the sensed indicator light; and When the autonomous navigation machine approaches an obstacle on the travel path, a spatial relationship between the autonomous navigation machine and the obstacle is obtained according to the at least one characteristic of the sensed indicator light; Among them, when the autonomous navigation machine gradually approaches a floating obstacle above the ground, the obstacle formed by the floating obstacle causes the light range of the vertical linear light to change, resulting in changes in the two line segments of the indicator light projected onto the ground and the floating obstacle, thereby judging that there is the obstacle ahead.

2. The obstacle detection method according to claim 1, characterized in that: The spatial relationship allows the autonomous navigation machine to calculate a distance between the autonomous navigation machine and the obstacle; after comparing the distance with a collision threshold, it is determined whether the autonomous navigation machine will collide with the obstacle; or, after comparing the distance with a drop threshold, it is determined whether the autonomous navigation machine will fall onto the obstacle.

3. The obstacle detection method according to claim 2, characterized in that: The vertical linear light is projected onto the ground and the obstacle; when the obstacle is a wall, the indicator light is sensed to form a first line segment and a second line segment, and the at least one feature is the length of the first line segment or the second line segment, the position of the second line segment, and / or the slope of the first line segment.

4. The obstacle detection method according to claim 2, characterized in that: The vertical linear light is projected onto the ground. When the obstacle is a cliff, the linear light is cut off by an edge of the ground. At least one characteristic of the indicator light is the length and / or slope of the indicator light projected onto the ground.

5. The obstacle detection method according to claim 2, characterized in that: The obstacle is a floating obstacle at a height from the ground. The indicator light projected on the ground and the obstacle forms a first line segment and a second line segment. At least one feature of the indicator light is the position of the second line segment.

6. An obstacle detection system, installed in an autonomous navigation machine, characterized in that: The system comprises: a controller; a light emitter coupled to the controller, for emitting an indication light projected onto a travel path of the autonomous navigation machine, wherein the indication light is a vertical linear light; a light sensor coupled to the controller and configured to sense the indicator light projected on the travel path, wherein the light emitter of the autonomous navigation machine is at a distance from the light sensor, wherein the light sensor is configured to sense at least one characteristic of the vertical linear light, namely, the length, position or slope of the vertical linear light; an image processor, coupled to the light sensor, for generating an image including the indicator light; A central processing unit is coupled to the image processor and the controller, and is used to execute an obstacle detection method, including: analyzing the image having the indicator light sensed by the optical sensor; Obtaining the at least one characteristic of the sensed indicator light; and When the autonomous navigation machine approaches an obstacle on the travel path, a spatial relationship between the autonomous navigation machine and the obstacle is obtained according to the at least one characteristic of the sensed indicator light; Among them, when the autonomous navigation machine gradually approaches a floating obstacle above the ground, the obstacle formed by the floating obstacle causes the light range of the vertical linear light to change, resulting in changes in the two line segments of the indicator light projected onto the ground and the floating obstacle, thereby judging that there is the obstacle ahead.

7. The obstacle detection system according to claim 6, characterized in that: The system obtains the spatial relationship so that the central processor calculates a distance between the autonomous navigation machine and the obstacle, and after comparing the distance with a collision threshold stored in a memory of the system, determines whether the autonomous navigation machine will collide with the obstacle threshold; Or after comparing the distance with a drop threshold stored in the memory, it is determined whether the autonomous navigation machine will fall onto the obstacle.

8. The obstacle detection system according to claim 7, characterized in that: The method executed by the central processor further includes instructing a controller to drive the autonomous navigation machine to avoid the obstacle when the autonomous navigation machine reaches the collision threshold or the drop threshold.

9. The obstacle detection system according to claim 7, characterized in that: The vertical linear light is projected onto the ground and the obstacle. When the obstacle is a wall, the indicator light is sensed to form a first line segment and a second line segment. The at least one feature is the length of the first line segment or the second line segment, the position of the second line segment, and / or the slope of the first line segment.

10. The obstacle detection system according to claim 7, characterized in that: The vertical linear light is projected onto the ground. When the obstacle is a cliff, the vertical linear light is cut off by an edge of the ground. At least one characteristic of the indicator light is the length and / or slope of the indicator light projected onto the ground.

11. The obstacle detection system according to claim 7, characterized in that: The obstacle is a floating obstacle at a height from the ground. The indicator light projected on the ground and the obstacle forms a first line segment and a second line segment. At least one feature of the indicator light is the position of the second line segment.

12. The obstacle detection system according to claim 11, characterized in that: The light emitter uses a laser or a light emitting diode as a light source to emit the vertical linear light as the indicating light.

13. A sensor subsystem for use in an autonomous navigation machine cruising in an area, characterized in that: The subsystem includes: A light emitter, used for emitting an indication light projected onto a travel path of the autonomous navigation machine, wherein the indication light is a vertical linear light; a light sensor for sensing the indicator light projected on the travel path, wherein the light emitter of the autonomous navigation machine is at a distance from the light sensor, wherein the light sensor is used to sense at least one characteristic of the vertical linear light, namely, the length, position or slope of the vertical linear light; and an image processor, coupled to the light sensor, for generating an image including the indicator light; When the autonomous navigation machine approaches an obstacle on the travel path, the autonomous navigation machine uses the at least one characteristic of the sensed indicator light to obtain a spatial relationship between the autonomous navigation machine and the obstacle; Among them, when the autonomous navigation machine gradually approaches a floating obstacle above the ground, the obstacle formed by the floating obstacle causes the light range of the vertical linear light to change, resulting in changes in the two line segments of the indicator light projected onto the ground and the floating obstacle, thereby judging that there is the obstacle ahead.

14. The sensor subsystem of claim 13, wherein: The subsystem is installed in an autonomous robot.

Images