KR20230093944A - Method for determining optimal driving path for autonomous mobile device and computing system performing the same - Google Patents

Abstract

Disclosed are a method for determining an optimal travel path of an autonomous mobile device capable of moving autonomously without human intervention and a computing system performing the same. According to one aspect of the present invention, as a method for determining an optimal driving path of an autonomous mobile device including a sensor for detecting a pedestrian, a computing system determines the autonomous mobile device based on data measured by the sensor. Generating a lattice-type probability map corresponding to a local area of the mobile device - wherein the lattice-type probability map is composed of unit lattices corresponding to each unit area obtained by dividing the local area into lattices of the same size, Assigning a probability value to each unit grid constituting the probability map - and determining, by the computing system, a driving route of the autonomous platform based on the grid-like probability map.

Description

Method for determining optimal driving path for autonomous mobile device and computing system performing the same}

The present invention relates to a method for determining an optimal travel path of an autonomous mobile device that can move autonomously without human intervention, such as an unmanned delivery robot, and a computing system for performing the same.

In order for an autonomous mobile device such as an unmanned delivery robot to drive autonomously, it must be clear about its surroundings, such as obstacles. Recognition of the surrounding environment is usually performed using various sensors, but there are obstacles that are difficult to recognize due to limitations. For example, water pits or very thin obstacles belong to this category. If an autonomous mobile device drives by itself in an environment with high uncertainty without recognizing these obstacles, it may lead to an unexpected accident. In order to solve this problem, it is necessary to recognize the pedestrian not only as an obstacle but also as an object of imitation from the robot's point of view. US Patent No. 10884417 may be cited as an example of an attempt to solve the problem in this respect. In US Patent No. 10884417, a specific pedestrian is followed in a complex environment with many pedestrians. However, when a specific pedestrian is followed, the direction in which the autonomous platform should proceed is not deeply considered, and thus, it is expected that there will be many unnecessary movements due to the pedestrian's moving path.

On the other hand, other conventional techniques include a method of generating a driving route using a map and a location estimation technique, a method of dividing a local area using a sensor and generating a driving route, and a method of generating a driving route using artificial intelligence. However, the previously acquired map-based driving route generation method has the disadvantage that it is difficult to change the driving route according to the changed environment, and the sensor-based local area estimation has the problem of high uncertainty due to the limitations of the sensor, and the unlearned uncertainty In high environments, artificial intelligence has limitations.

US Patent No. 10884417

The present invention has been made to solve these problems, and the technical problem to be achieved by the present invention is to enable an autonomous mobile device such as an unmanned robot to select and move a safer driving route in an evil situation that is difficult to recognize using a sensor. is to provide a method for

In addition, the problem to be solved by the present invention is to provide a method of predicting a driving path on which a robot can stably drive by probabilistically analyzing a moving path of the public. Since the human cognitive function is capable of recognizing even obstacles that are difficult for a robot to recognize using a sensor, the present invention uses the excellent cognitive ability of a human to imitate the movement path of a number of pedestrians, thereby optimizing the autonomous mobile device's optimal driving path. suggests a way to determine

According to one aspect of the present invention, as a method for determining an optimal driving path of an autonomous mobile device including a sensor for detecting a pedestrian, a computing system determines the autonomous mobile device based on data measured by the sensor. Generating a lattice-type probability map corresponding to a local area of the mobile device - wherein the lattice-type probability map is composed of unit lattices corresponding to each unit area obtained by dividing the local area into lattices of the same size, Assigning a probability value to each unit grid constituting the probability map - and determining, by the computing system, a driving route of the autonomous platform based on the grid-like probability map.

In an embodiment, the probability value assigned to each unit grid constituting the grid-like probability map may increase as the number of pedestrians visiting the unit area on the local area corresponding to the unit cell increases.

In an embodiment, generating a grid-like probability map corresponding to a local area of the autonomous platform based on data measured by the sensor may include: based on the data measured by the sensor, the local area of the autonomous platform Detecting a pedestrian moving within and tracking the movement of the pedestrian, updating, by the computing system, a grid-type information map corresponding to the local area based on a tracking result of the movement of the pedestrian - wherein , The grid-type information map is composed of information grids corresponding to each unit area constituting the local area, and each information grid constituting the grid-type information map includes a unit on the local area corresponding to the information grid. Assigning a numerical value determined based on data measured by the sensor to pedestrians visiting the area - and generating, by the computing system, the grid-like probability map based on the grid-like information map can do.

In an embodiment, the updating of the grid-type information map corresponding to the local area based on the tracking result of the pedestrian may include: whenever the pedestrian visits a specific unit area constituting the local area, the pedestrian and accumulating weights in the information grid corresponding to the visited specific unit area.

In one embodiment, the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian may be calculated based on the movement speed of the pedestrian when passing the specific unit area.

In one embodiment, the generating of the grid-like probability map based on the grid-like information map may include performing a convolution operation using a Gaussian filter whose center value is normalized to 1 on the grid-like information map to obtain the grid It may include generating a type probability map.

In one embodiment, the step of determining the driving route of the autonomous platform based on the grid-like probability map may include forming a graph based on the grid-like probability map, a unit where the starting point of the autonomous platform is located. detecting a shortest path between a vertex corresponding to an area and a vertex corresponding to a unit area where a destination of the autonomous platform is located, and determining a travel path of the autonomous platform based on the shortest path, The forming of the graph may include generating a vertex corresponding to each of the unit lattices constituting the lattice-like probability map. Generating an edge connecting two vertices corresponding to two unit lattices, and assigning a weight of each edge included in the graph to two unit lattices corresponding to each of the two vertices connected by the edge. It may include determining based on the value.

According to another aspect of the present invention, a recorded computer program installed in a data processing device and performing the above method is provided.

According to another aspect of the present invention, a computing system includes a processor and a memory for storing a computer program, wherein the computer program, when executed by the processor, causes the computing system to perform the above-described method. A computing system is provided.

According to another aspect of the present invention, a computing system for determining an optimal driving path of an autonomous platform including a sensor for detecting a pedestrian, corresponding to a local area of the autonomous platform based on data measured by the sensor. A generation module for generating a lattice-type probability map, wherein the lattice-type probability map is composed of unit lattices corresponding to respective unit areas obtained by dividing the local area into lattices of the same size, and each of the lattice-type probability maps constituting the lattice-type probability map Probability values are assigned to unit grids of - and a determination module for determining a driving route of the autonomous platform based on the grid-like probability map is provided.

In one embodiment, the generation module detects a pedestrian moving within a local area of the autonomous mobile device based on data measured by the sensor, and a tracking module for tracking the movement of the pedestrian, A grid information map generation module for updating a grid information map corresponding to the local area based on a tracking result for the local area, wherein the grid information map is composed of an information grid corresponding to each unit area constituting the local area; , Each information grid constituting the grid-type information map is assigned a numerical value determined based on data measured by the sensor for pedestrians who have visited a unit area on the local area corresponding to the information grid. - and a lattice probability map generating module for generating the lattice-like probability map based on the lattice-like information map.

In an embodiment, the grid information map generation module may accumulate weights in an information grid corresponding to the specific unit area visited by the pedestrian whenever the pedestrian visits a specific unit area constituting the local area. .

In an embodiment, the grid information map generating module calculates the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian based on the moving speed of the pedestrian when passing the specific unit area. can

In an embodiment, the lattice probability map generating module may generate the lattice-type probability map by performing a convolution operation using a Gaussian filter whose center value is normalized to 1 on the lattice-type information map.

In one embodiment, the determination module may include a graph module for forming a graph based on the lattice-like probability map and a vertex corresponding to a unit area where the starting point of the autonomous platform is located and a destination of the autonomous platform are located. A driving path determining module for detecting a shortest path between vertices corresponding to a unit area and determining a driving path of the autonomous platform based on the shortest path, wherein the graph module is a unit constituting the lattice-type probability map. Vertices corresponding to each of the lattices are generated, and when probability values assigned to two adjacent unit lattices are all equal to or greater than a predetermined threshold value, an edge connecting the two vertexes corresponding to the two adjacent unit lattices is generated; The weight of each edge included in the graph may be determined based on probability values assigned to two unit cells corresponding to each of the two vertices connected by the edge.

According to the technical idea of the present invention, it is possible to provide a method for determining an optimal travel path of an autonomous mobile device that can move autonomously without human intervention, such as an unmanned delivery robot.

According to the technical idea of the present invention, it is possible to provide a method of predicting a driving path on which a robot can stably travel by probabilistically analyzing a moving path of the public. Since the human cognitive function is capable of recognizing even obstacles that are difficult for a robot to recognize using a sensor, the present invention uses the excellent cognitive ability of a human to imitate the movement path of a number of pedestrians, thereby optimizing the autonomous mobile device's optimal driving path. can decide

In order to more fully understand the drawings cited in the detailed description of the present invention, a brief description of each drawing is provided.
FIG. 1 is a diagram schematically illustrating an environment in which a schematic configuration of an autonomous platform and a method for determining an optimal driving path of an autonomous platform according to an embodiment of the present invention are performed.
2 is a flowchart schematically illustrating a method for determining an optimal driving path according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating an example of step S100 of FIG. 2 in more detail.
4A is a diagram illustrating an example of a local area and a unit area.
FIG. 4B is a diagram illustrating an example of a grid-type information map corresponding to the local area of FIG. 4A.
5 is a diagram showing a lattice-like probability map in the form of a heat map.
6 is a flowchart illustrating an example of a specific method for determining a travel path of an autonomous platform, and FIGS. 7A to 7D are diagrams for explaining a process of generating a graph by the method of FIG. 6 .
8 is a block diagram showing a schematic configuration of a computing system according to an embodiment of the present invention.

Since the present invention can apply various transformations and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. However, it should be understood that this is not intended to limit the present invention to specific embodiments, and includes all transformations, equivalents, and substitutes included in the spirit and scope of the present invention. In describing the present invention, if it is determined that a detailed description of related known technologies may obscure the gist of the present invention, the detailed description will be omitted.

Terms such as first and second may be used to describe various components, but the components should not be limited by the terms. These terms are only used for the purpose of distinguishing one component from another.

Terms used in this application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise.

In this specification, terms such as "include" or "have" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other It should be understood that the presence or addition of features, numbers, steps, operations, components, parts, or combinations thereof is not precluded.

In addition, in the present specification, when one component 'transmits' data to another component, the component may directly transmit the data to the other component, or through at least one other component. It means that the data can be transmitted to the other component. Conversely, when one component 'directly transmits' data to another component, it means that the data is transmitted from the component to the other component without going through the other component.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings, focusing on embodiments of the present invention. Like reference numerals in each figure indicate like elements.

1 is a diagram schematically showing an environment in which a schematic configuration of an autonomous platform and a method for determining an optimal travel path of the autonomous platform according to an embodiment of the present invention are performed. The autonomous mobile device 200 is a device capable of driving automatically without a human driving, and may include an unmanned robot, an unmanned car, and the like.

Referring to FIG. 1 , the autonomous mobile device 200 may include a sensor 210 and a computing system 100 .

The sensor 210 may be a sensor for measuring data used for detecting an object (in particular, a pedestrian). For example, the sensor 210 may be a camera sensor, a vision sensor, a lidar sensor, a radar sensor, and the like.

The computing system 100 provides a method for determining an optimal driving route of the autonomous platform 100 based on data detected by the sensor 210 (hereinafter, referred to as 'optimal driving route determination method'). can be done The computing system 100 may use the sensor 210 of the autonomous mobile device 200 to recognize nearby pedestrians, track the pedestrian's movement path, and accumulate the pedestrian's movement path to obtain a stochastic path. You can create maps. In addition, a local driving route of the autonomous platform 100 may be estimated using the probabilistic route map generated in this way. Meanwhile, according to an embodiment, the computing system 100 may estimate the speed of the pedestrian and use the estimated speed as a weight when generating a probabilistic path map.

The computing system 100 may be an information processing device. The computing system 100 may be an embedded system, but is not limited thereto. The computing system 100 may be a processing system such as a computer, a mobile phone, a satellite phone, a wireless phone, a Session Initiation Protocol (SIP), a Wireless Local Loop (WLL) station, a smart phone, a tablet PC, and a personal digital assistant (PDA). It may also be a processing device including a handheld device such as an assistant.

Meanwhile, FIG. 1 shows an example in which the computing system 100 is included inside the autonomous platform 200, but according to an embodiment, the computing system 100 is outside the autonomous platform 200. In this case, the computing system 100 may receive the data measured by the sensor 210 through a wireless network and perform an optimal driving path determination method according to the technical concept of the present invention.

Meanwhile, as shown in FIG. 1 , the computing system 100 may be connected to a predetermined remote server 300 through a wired/wireless network. In FIG. 1 , the remote server 300 is shown as being connected to one computing system 100, but in reality, it may be connected to a plurality of computing systems included in a plurality of autonomous mobile devices through a wired/wireless network.

The remote server 300 may estimate a global driving route by integrating optimal local driving routes determined by the plurality of computing systems 100 . Alternatively, the remote server 300 estimates a global driving path by receiving data measured by each sensor included in a plurality of autonomous mobility devices and directly performing an optimal driving path determining method according to the technical concept of the present invention based on the received data. can do.

2 is a flowchart schematically illustrating a method for determining an optimal driving path according to an embodiment of the present invention.

Referring to FIG. 2 , the computing system 100 may generate a grid-type probability map corresponding to a local area of the autonomous platform 200 based on data measured by the sensor 210 (S100). .

The local area may be a predetermined specific area or an area of a certain size centered on the current location of the autonomous platform 200 . For example, the local area may be a coverage area capable of detecting or tracking a pedestrian through the sensor 210 . Alternatively, the local area may be a drivable area in which the autonomous platform 200 can drive. Alternatively, the local area may be the inside of a specific building or a specific area inside a building.

The local area may be composed of unit areas obtained by dividing the local area into lattices having the same size, and the lattice-like probability map is composed of unit lattices corresponding to each of the unit areas obtained by dividing the local area into lattices having the same size. can be configured.

A probability value may be assigned to each unit cell constituting the grid-like probability map. A probability value assigned to each unit grid constituting the grid-like probability map may be determined based on the number of pedestrians visiting the unit area corresponding to the unit grid and/or the speed of the visiting pedestrians. For example, in an embodiment, the probability value assigned to each unit cell constituting the grid-like probability map may increase as the number of pedestrians visiting the unit area on the local area corresponding to the unit cell increases.

FIG. 3 is a flowchart illustrating an example of step S100 of FIG. 2 in more detail.

Referring to FIG. 3 , the computing system 100 first generates a lattice-type probability map corresponding to a local area of the autonomous platform 200 based on data measured by the sensor 210. Based on the data measured in step 210, a pedestrian moving within the local area of the autonomous platform 200 may be detected, and the movement of the pedestrian may be tracked (S110).

A method of detecting a pedestrian through data measured through a sensor such as a vision sensor, lidar, or radar is widely known, and a specific method of detecting a pedestrian through data measured through a sensor is a key technical feature of the present invention. Since it is not, detailed description will be omitted in this specification.

Meanwhile, the computing system 100 may update a grid-type information map corresponding to the local area based on a tracking result of the pedestrian's movement (S120).

At this time, the grid-type information map is composed of information grids corresponding to each unit area constituting the local area, and each information grid constituting the grid-type information map includes a unit on the local area corresponding to the information grid. A numerical value determined based on data measured by the sensor for pedestrians visiting the area may be assigned.

4A is a diagram illustrating an example of the aforementioned local area and unit area.

The local area 10 shown in FIG. 4A may be divided into grid-shaped unit areas (for example, 11), and in the example of FIG. 4A, the case where the local area 10 is divided into a grid of 8×10 are showing However, it is common in the field to which the present invention belongs that the line for division is not actually displayed in the local area 10 as shown in FIG. 4A, and the division line shown in FIG. 4A is only for convenience of understanding. Anyone who has knowledge of this will be able to understand clearly.

Meanwhile, obstacles may exist in the local area 10 . In Fig. 4a, obstacles are indicated by black boxes. An obstacle may refer to an object that hinders pedestrians and the autonomous platform 10 from moving, and is smaller than the resolution of the sensor 210, such as a stretched line, or difficult to distinguish from the ground through the sensor 210, such as muddy water. When the path proceeds, obstacles that may cause obstacles to the driving of the autonomous platform may be included.

Meanwhile, in FIG. 4A , a pedestrian and a movement path of the pedestrian are shown together. As described above, the computing system 100 may detect a pedestrian and track the movement path of the detected pedestrian.

FIG. 4B is a diagram illustrating an example of a grid-type information map corresponding to the local area of FIG. 4A.

As described above, the grid-type information map 20 may be composed of information grids (for example, 21) corresponding to each unit area constituting the local area, and the grid-type information map 20 of FIG. 4B is composed of an 8×10 information grid.

Also, as described above, the computing system 100 may update the grid-type information map 20 corresponding to the local area 10 based on the tracking result of the pedestrian's movement, as shown in FIG. 4A. As described above, when the pedestrian moves, the computing system 10 assigns a predetermined weight to the information grid (for example, the movement speed of the pedestrian) whenever the pedestrian visits a unit area corresponding to a specific information grid; FIG. 4B Assuming that all moving speeds are 1 in , the grid-type information map 20 may be updated as shown in FIG. 4B by accumulating the movement speeds.

Referring back to FIG. 3 , the computing system 100 may generate the lattice-type probability map based on the lattice-type information map (S130).

In an embodiment, the computing system 100 may generate the grid-like probability map by performing a convolution operation using a Gaussian filter whose center value is normalized to 1 on the grid-like information map.

In addition to this, of course, there may be various methods for generating a grid-like probability map from a grid-like information map.

Meanwhile, depending on the embodiment, in order to prevent only a specific unit area or a specific path from having a high probability value, a threshold value may be assigned when a value accumulated in an information grid corresponding to a specific unit area exceeds a predetermined threshold. there is.

5 is a diagram showing a lattice-like probability map in the form of a heat map. For convenience of understanding, in this specification, the lattice-type probability map is shown in the form of a heat map. On the other hand, each unit grid of the grid-like probability map (heat map) of FIG. 5 corresponds to a unit area, and FIG. 5 shows a grid-like probability map (heat map) corresponding to a local area divided into 15*15. . In FIG. 5 , a light-colored lattice has a high probability value and a dark-colored lattice has a low probability value.

Meanwhile, in an embodiment, the computing system 100 is a method of accumulating weights in an information grid corresponding to the specific unit area visited by the pedestrian whenever the pedestrian visits a specific unit area constituting the local area. A grid-type information map corresponding to the local area can be updated with In this way, in the present invention, as the number of pedestrians visiting a specific unit area increases, a higher value can be assigned to the information grid corresponding to the unit area, and a higher probability value can be assigned to the unit grid corresponding to the unit area. there is.

Also, in an embodiment, the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian may be calculated based on a moving speed of the pedestrian when passing the specific unit area. The higher the speed of the pedestrian passing through a specific unit area, the higher the possibility that the corresponding unit area is an area that is efficient for traffic. Therefore, by doing this, in the present invention, as the walking speed of a pedestrian visiting a specific unit area increases, a higher value can be assigned to the information grid corresponding to the unit area, and consequently, a higher probability to the unit grid corresponding to the unit area. A value can be assigned.

Meanwhile, the computing system 100 may determine a driving path of the autonomous platform 200 based on the generated grid-like probability map.

The computing system 100 may determine the driving path of the autonomous platform 200 from the lattice-like probability map using various methods, and as one example of these, the computing system 100 may use a graph. . 6 is a flowchart illustrating an example of a specific method for the computing system 100 to determine a driving path of the autonomous platform 200 using a graph according to an embodiment of the present invention, and FIGS. 7A to 7D is a diagram for explaining a process of generating a graph by the method of FIG. 6 .

Referring to FIG. 6 , the computing system 100 may form a graph (in particular, an undirected graph) based on the grid-like probability map (S210).

More specifically, in order to form a graph, the computing system 100 may generate a vertex corresponding to each of the unit lattices to which a probability value equal to or greater than a predetermined threshold is assigned among unit lattices constituting the lattice-like probability map ( S211). If the lattice-type probability map is as shown in FIG. 7A and the threshold value is 0.02, the computing system 100 corresponds to each of the unit lattices constituting the lattice-type probability map to which a probability value of 0.02 or more is assigned. A vertex can be generated as shown in FIG. 7B. For reference, in each vertex of FIG. 7B, a probability value assigned to a corresponding unit cell is described.

Meanwhile, the computing system 100 may generate an edge connecting two vertices corresponding to two adjacent unit cells (S212). FIG. 7C shows a form in which trunk lines are connected to the vertices of FIG. 7B. In the example of FIG. 7C , the edges are connected only when the unit cells are vertically adjacent to each other, but according to embodiments, the computing system 100 may connect the edges even when the unit cells are diagonally adjacent to each other.

Then, the computing system 100 may determine the weight of each edge included in the graph based on the probability values assigned to the two unit cells corresponding to each of the two vertices connected by the edge (S213). For example, the computing system 100 may determine the weight of an edge as the reciprocal of the sum of probability values assigned to two unit cells corresponding to each of the two vertices connected by the corresponding edge. The result of weighting the edges of Fig. 7c is shown in Fig. 7d.

를 곱한 값으로 결정할 수도 있다.On the other hand, in the case of an embodiment in which edges are connected even when two unit cells are diagonally adjacent, the computing system 100 determines that the weight of the edge connecting vertices adjacent to each other vertically and horizontally corresponds to each of the two vertices connected by the corresponding edge. It is determined by the reciprocal of the sum of the probability values assigned to the two unit lattices, and the weight of the edge connecting diagonally adjacent vertices is the sum of the probability values assigned to the two unit lattices corresponding to each of the two vertices connected by the corresponding edge. in reciprocal

It can also be determined as a value multiplied by .

로 설정할 수도 있다.In the previous example, the computing system 100 used a method of determining weights using probability values assigned to each unit cell, but the technical idea of the present invention is not limited to this, and depending on the embodiment, all of the edges The same weight can be set, and the weight of the edge connecting vertices adjacent to the top, bottom, left, and right is set to 1, and the weight of the edge connecting diagonally adjacent vertices is

can also be set to

Referring back to FIG. 6 , after a graph is created, the computing system 100 uses the graph to determine a vertex corresponding to the unit area where the starting point of the autonomous platform is located and the destination of the autonomous platform 200. A shortest path between vertices corresponding to the located unit area may be detected (S220). In one embodiment, the computing system 100 may determine the shortest path using a Dijkstra algorithm, a Floyd-Warshall algorithm, or the like.

Thereafter, the computing system 100 may determine a driving path of the autonomous platform 200 based on the determined shortest path (S230). For example, the computing system 100 may determine the driving path of the autonomous platform 200 to include a unit area corresponding to each vertex included in the shortest path.

8 is a block diagram showing a schematic configuration of a computing system 100 according to an embodiment of the present invention.

Referring to FIG. 8 , the computing system 100 may include a generation module 110 and a determination module 120. Depending on embodiments, the computing system 100 may include only some of the components shown in FIG. 8 or may include more components than these. For example, the computing system 100 may control at least one communication module capable of communicating with an external system through a network or functions and/or resources of other components included in the computing system 100. It may further include a control module, a storage module capable of storing various types of information, an input/output module that interfaces with a user and receives information from the outside or outputs information to the outside.

The computing system 100 may include hardware resources and/or software required to implement the technical idea of the present invention, and does not necessarily mean a single physical component or a single device. . That is, the computing system 100 may refer to a logical combination of hardware and/or software provided to implement the technical idea of the present invention, and if necessary, is installed in devices spaced apart from each other to perform respective functions. By doing so, it may be implemented as a set of logical configurations for implementing the technical idea of the present invention. In addition, the computing system 100 may refer to a set of components implemented separately for each function or role to implement the technical idea of the present invention.

In this specification, a module may mean a functional and structural combination of hardware for implementing the technical idea of the present invention and software for driving the hardware. For example, the module may refer to a predetermined code and a logical unit of hardware resources for executing the predetermined code, and does not necessarily mean a physically connected code or one type of hardware according to the present invention. can be easily deduced to the average expert in the art.

The generation module 110 may generate a lattice-type probability map corresponding to a local area of the autonomous platform 200 based on data measured by the sensor 210 embedded in the autonomous platform 200 . Here, the lattice-type probability map is composed of unit lattices corresponding to each unit area obtained by dividing the local area into lattices of the same size, and a probability value may be assigned to each unit lattice constituting the lattice-type probability map. .

The generation module 110 may include a tracking module 111, a grid information map generation module 112, and a grid probability map generation module 113.

The tracking module 111 may detect a pedestrian moving within a local area of the autonomous platform 200 based on data measured by the sensor 210 and track the movement of the pedestrian.

The grid information map generating module 112 may update a grid-type information map corresponding to the local area based on a tracking result of the pedestrian's movement, wherein the grid-type information map constitutes the local area. In each information grid constituting the grid-type information map, the sensor measures a pedestrian visiting a unit area on the local area corresponding to the information grid. A numerical value determined based on the data may be assigned.

In one embodiment, the grid information map generation module 112 is configured to accumulate weights in the information grid corresponding to the specific unit area visited by the pedestrian whenever the pedestrian visits a specific unit area constituting the local area. can

In addition, according to an embodiment, the grid information map generation module 112 determines the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian based on the movement speed when the pedestrian passes the specific unit area. can be calculated

The lattice probability map generation module 113 may generate the lattice-type probability map based on the lattice-type information map.

In one embodiment, the lattice probability map generation module 113 may generate the lattice-type probability map by performing a convolution operation using a Gaussian filter whose center value is normalized to 1 on the lattice-type information map.

Meanwhile, the determination module 120 may determine a driving path of the autonomous platform 200 based on the lattice-like probability map.

The determination module 120 may include a graph module 121 and a driving path determination module 122 .

The graph module 121 may form a graph based on the lattice-like probability map, and more specifically, generate a vertex corresponding to each unit lattice constituting the lattice-like probability map, and create two units adjacent to each other. When all probability values assigned to the lattice are equal to or greater than a predetermined threshold value, an edge connecting two vertices corresponding to the two adjacent unit lattices is generated, and the weight of each edge included in the graph is connected by the edge. It can be determined based on probability values assigned to two unit cells corresponding to each of the two vertices.

The driving route determination module 122 detects the shortest path between a vertex corresponding to the unit area where the starting point of the autonomous platform is located and a vertex corresponding to the unit area where the destination of the autonomous platform is located, and the shortest path Based on this, it is possible to determine a driving path of the autonomous platform.

Meanwhile, according to an implementation example, the computing system 100 may include at least one processor and a memory storing a program executed by the processor. The processor may include a single-core CPU or a multi-core CPU. The memory may include high-speed random access memory and may also include non-volatile memory such as one or more magnetic disk storage devices, flash memory devices, or other non-volatile solid-state memory devices. Access to memory by processors and other components may be controlled by a memory controller.

Meanwhile, the method for measuring the length of a living tissue/lesional tissue according to an embodiment of the present invention may be implemented in the form of computer-readable program instructions and stored in a computer-readable recording medium. The computer-readable recording medium includes all types of recording devices in which data that can be read by a computer system is stored.

Program commands recorded on the recording medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in the software field.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, floptical disks and Hardware devices specially configured to store and execute program instructions, such as magneto-optical media and ROM, RAM, flash memory, and the like. In addition, the computer-readable recording medium is distributed in computer systems connected through a network, so that computer-readable codes can be stored and executed in a distributed manner.

Examples of program instructions include high-level language codes that can be executed by a device that electronically processes information using an interpreter, for example, a computer, as well as machine language codes generated by a compiler.

The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

The above description of the present invention is for illustrative purposes, and those skilled in the art can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts thereof should be construed as being included in the scope of the present invention. .

Claims (16)

As a method for determining an optimal driving path of an autonomous mobile device including a sensor for detecting a pedestrian,
Generating, by a computing system, a lattice-like probability map corresponding to a local area of the autonomous platform based on data measured by the sensor, wherein the lattice-like probability map divides the local area into grids of the same size. a unit cell corresponding to each unit area, and a probability value is assigned to each unit cell constituting the grid-like probability map; and
and determining, by the computing system, a driving path of the autonomous platform based on the grid-like probability map.
According to claim 1,
The probability value assigned to each unit cell constituting the grid-like probability map is
The method characterized in that the size increases as the number of pedestrians visiting the unit area on the local area corresponding to the unit grid increases.
According to claim 1,
Generating a lattice-like probability map corresponding to a local area of the autonomous platform based on the data measured by the sensor,
detecting a pedestrian moving within a local area of the autonomous platform based on data measured by the sensor, and tracking the movement of the pedestrian;
Updating, by the computing system, a lattice-type information map corresponding to the local area based on a tracking result of the pedestrian's movement, wherein the lattice-type information map corresponds to each unit area constituting the local area. It consists of an information grid that
Each information grid constituting the grid-type information map is assigned a numerical value determined based on data measured by the sensor for pedestrians who have visited a unit area on the local area corresponding to the information grid; and
generating, by the computing system, the grid-like probability map based on the grid-like information map.
According to claim 3,
Updating the grid-type information map corresponding to the local area based on the tracking result of the pedestrian,
and accumulating weights in an information grid corresponding to the specific unit area visited by the pedestrian whenever the pedestrian visits a specific unit area constituting the local area.
According to claim 4,
The method of claim 1 , wherein the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian is calculated based on a moving speed of the pedestrian when passing the specific unit area.
According to claim 3,
Generating the grid-like probability map based on the grid-like information map comprises:
and generating the grid-like probability map by performing a convolution operation on the grid-like information map using a Gaussian filter whose center value is normalized to 1.
According to claim 1,
Determining a driving path of the autonomous platform based on the grid-like probability map includes:
forming a graph based on the lattice-like probability map;
detecting a shortest path between a vertex corresponding to a unit area where a starting point of the autonomous platform is located and a vertex corresponding to a unit area where a destination of the autonomous platform is located; and
Determining a driving path of the autonomous platform based on the shortest path;
Forming the graph is
generating a vertex corresponding to each unit cell constituting the grid-type probability map;
generating an edge connecting two vertices corresponding to the two adjacent unit cells when all probability values assigned to the two unit cells adjacent to each other are equal to or greater than a predetermined threshold value; and
and determining a weight of each edge included in the graph based on a probability value assigned to two unit cells corresponding to each of two vertices connected by the edge.
A recorded computer program installed in a data processing device and for carrying out the method according to claims 1 to 7 above.
As a computing system,
processor; and
a memory for storing a computer program;
The computer program, when executed by the processor, causes the computing system to perform the method according to any one of claims 1 to 7.
A computing system for determining an optimal driving path of an autonomous mobile device including a sensor for detecting a pedestrian,
A generation module for generating a lattice-like probability map corresponding to the local area of the autonomous platform based on the data measured by the sensor, wherein the lattice-like probability map is each unit area obtained by dividing the local area into grids of the same size. It is composed of a unit cell corresponding to , and a probability value is assigned to each unit cell constituting the grid-like probability map; and
and a determination module for determining a driving path of the autonomous platform based on the lattice-like probability map.
According to claim 10,
The probability value assigned to each unit cell constituting the grid-like probability map is
The computing system, characterized in that the size increases as the number of pedestrians visiting the unit area on the local area corresponding to the unit grid increases.
According to claim 10,
The generating module,
a tracking module that detects a pedestrian moving within a local area of the autonomous platform based on the data measured by the sensor and tracks the movement of the pedestrian;
A grid information map generation module for updating a grid information map corresponding to the local area based on a tracking result of the movement of the pedestrian, wherein the grid information map corresponds to each unit area constituting the local area It consists of an information grid,
Each information grid constituting the grid-type information map is assigned a numerical value determined based on data measured by the sensor for pedestrians who have visited a unit area on the local area corresponding to the information grid; and
and a lattice probability map generating module for generating the lattice-like probability map based on the lattice-like information map.
According to claim 12,
The grid information map generation module,
Whenever the pedestrian visits a specific unit area constituting the local area, a weight is accumulated in an information grid corresponding to the specific unit area visited by the pedestrian.
According to claim 13,
The grid information map generation module,
The computing system for calculating the weight accumulated in the information grid corresponding to the specific unit area visited by the pedestrian based on the moving speed of the pedestrian when passing the specific unit area.
According to claim 12,
The lattice probability map generation module,
A computing system for generating the grid-like probability map by performing a convolution operation using a Gaussian filter whose center value is normalized to 1 on the grid-like information map.
According to claim 10,
The decision module,
a graph module for forming a graph based on the lattice-like probability map; and
A shortest path between a vertex corresponding to the unit area where the starting point of the autonomous platform is located and a vertex corresponding to the unit area where the destination of the autonomous platform is located is detected, and the autonomous platform travels based on the shortest path. Including a driving route determination module for determining a route,
The graph module,
Generating vertices corresponding to each unit lattice constituting the lattice-like probability map;
When all probability values assigned to two unit cells adjacent to each other are equal to or greater than a predetermined threshold value, an edge connecting two vertices corresponding to the two adjacent unit cells is generated;
A computing system for determining weights of each edge included in the graph based on probability values assigned to two unit cells corresponding to two vertices connected by the edge.

Images