JP4953313B2 - Environment recognition system, autonomous mobile body and environment recognition program - Google Patents

Description

  The present invention relates to an autonomous mobile body, an environment recognition system, and an environment recognition program that recognize a movable route in an environment where an object such as an obstacle exists.

  In recent years, the usefulness of intelligent mobile bodies such as robots has been extended not only to automation and unmanned production of living industries, but also to life support for general households, and research and development have been actively conducted. The robot needs to be able to act autonomously in order for the robot to support life and the like, and in order for the robot to act autonomously, the robot needs to recognize the surrounding environment. For example, it is necessary to constantly acquire a range in which the robot can act, that is, a so-called open space.

In order for the robot to recognize the surrounding environment, a laser range finder (LRF) as a two-dimensional range sensor is particularly useful. Since the laser range finder can measure the distance to an obstacle or a moving body accurately and at high speed by measuring the point where the laser beam hit with a camera, for example, the robot avoids the obstacle, It can be used for the purpose of following the moving body. In addition, since the laser range finder has a relatively narrow range in which laser can be measured, a technique for using in combination with other sensors is also known in order to target distant obstacles.
As a technique for recognizing the surrounding environment by combining a laser range finder and other sensors, for example, the following conventional technique (J01) is known.

(J01) Technology described in Patent Document 1 (Japanese Patent Laid-Open No. 2000-329852) Patent Document 1 discloses a prior art obtained by performing stereo image processing on images captured by two electronic cameras (1, 2). A measured value of the distance to the obstacle such as a vehicle and a measured value of the distance to the obstacle obtained by the laser range finder (7) are measured, and the obstacle is based on the measured values of two distances. The technology about the obstacle recognition device of the car which recognizes is described.

JP 2000-329852 A (“0025”, “0068” to “0079” FIGS. 1 to 11)

(Problems of conventional technology)
The prior art (J01) requires expensive cameras 1 and 2 to be arranged with high precision in order to measure the distance to an obstacle by stereo image processing. In addition, the stereo image analysis has a problem that the amount of calculation for measuring the distance becomes large. Therefore, the prior art (J01) has a problem that costs such as cost and calculation amount increase. In particular, there is a problem that it is difficult to apply to an intelligent moving body such as an autonomous behavior type robot.

In view of the above circumstances, the present invention has the following description (O01) as a technical problem.
(O01) To recognize the surrounding environment at high speed and with high accuracy and reduce the cost.

In order to solve the technical problem, an environment recognition system according to claim 1 comprises:
A distance measuring device for detecting an object existing in the distance measuring area by measuring a distance with respect to a predetermined distance measuring area;
An imaging device that captures an image including color information with respect to an imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object non-existing space area in which no object exists in the ranging area based on a measurement result by the ranging device;
Based on the image captured by the imaging device, spatial color information acquisition means for acquiring spatial color information that is color information of the object non-existing spatial region and an overlapping imaging region;
Space discriminating means for discriminating an area of color information matching the spatial color information in the imaging area as an object non-existing space based on an image captured by the imaging apparatus;
It is provided with.

The invention according to claim 2 is the environment recognition system according to claim 1,
Based on the image captured by the imaging device, when the spatial color information matches the color information of the area adjacent to the area of the spatial color information in the imaging area, The space discriminating means for discriminating that the space exists;
It is provided with.

The invention according to claim 3 is the environment recognition system according to claim 1 or 2,
Based on the image captured by the imaging device, the spatial color information acquisition unit that acquires, as the spatial color information, an average value of the color information of the object non-existing space region and the overlapping imaging region;
Based on the image captured by the imaging device, the color information matches the spatial color information when there is color information in the imaging region that is less than a preset threshold value with respect to the spatial color information. The space discriminating means which regards the area of the color information as the color information and discriminates the area of the color information as the object non-existing space;
It is provided with.

In order to solve the technical problem, an environment recognition system according to claim 4 is provided.
A distance measuring device for detecting an object existing in the distance measuring area by measuring a distance with respect to a predetermined distance measuring area;
An imaging device that captures an image including color information with respect to an imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object existence space area where an object exists in the ranging area based on a measurement result by the ranging device;
Spatial color information acquisition means for acquiring spatial color information that is color information of the imaging region that overlaps the object presence space region based on an image captured by the imaging device;
Space discriminating means for discriminating, in the imaging area, an area of color information that matches the spatial color information as an object existence space based on an image captured by the imaging apparatus;
It is provided with.

The invention according to claim 5 is the environment recognition system according to any one of claims 1 to 4,
The distance measuring device configured by a laser range finder that measures a distance to an object existing in the distance measuring area by scanning laser light in the distance measuring area and detecting laser light reflected by the object. When,
It is provided with.

In order to solve the technical problem, the autonomous mobile body of the invention according to claim 6 is:
A distance measuring device that detects an object existing in the distance measuring area by measuring a distance with respect to a distance measuring area set in front of the moving body;
An imaging device that captures an image including color information with respect to a front imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object non-existing space area in which no object exists in the ranging area based on a measurement result by the ranging device;
Based on the image captured by the imaging device, spatial color information acquisition means for acquiring spatial color information that is color information of the object non-existing spatial region and an overlapping imaging region;
Space discriminating means for discriminating an area of color information matching the spatial color information in the imaging area as an object non-existing space based on an image captured by the imaging apparatus;
It is determined that the object non-existing space is a movable area, and a route setting means for setting a moving route that avoids the object;
Movement control means for moving the moving body along the set movement path;
It is provided with.

In order to solve the technical problem, the environment recognition program of the invention according to claim 7
Computer
Based on the measurement result by the distance measuring device that detects the object existing in the distance measurement area by measuring the distance with respect to the predetermined distance measurement area, no object is not present in the distance measurement area. Ranging area space specifying means for specifying a spatial area,
The object non-existing space is based on an image captured by an imaging device that captures an image including color information for an imaging region that overlaps a ranging region measured by the ranging device and is wider than the ranging region. Spatial color information acquisition means for acquiring spatial color information which is color information of an area and an overlapping imaging area;
Space discriminating means for discriminating, based on an image captured by the imaging device, an area of color information that matches the spatial color information in the imaging area as an object non-existing space in which no object exists;
To function as.

  According to the first aspect of the present invention, an object non-existing space area in the ranging area is specified based on a measurement result by the ranging apparatus, and based on an image captured by the imaging apparatus, Spatial color information of the imaging region that overlaps the object non-existing space region is acquired, and it is determined that the color information region that matches the spatial color information is the object non-existing space of the imaging region. As a result, the object non-existing space can be expanded from the object non-existing space area in the ranging area to the imaging area, and an open space where no object exists can be recognized accurately and over a wide range. it can. In addition, the cost of the imaging device, the amount of calculation, and the like can be reduced as compared with the prior art (J01) that performs stereo image analysis using two or more imaging devices.

According to the second aspect of the present invention, since it is not determined whether or not the area that is not adjacent to the area of the spatial color information is the object non-existing space as compared with the case without the configuration of the present invention, The amount of computation for determining the object non-existing space can be reduced, and for example, the object is not determined for a closed space such as an object non-existing space surrounded by the object, and only the region adjacent to the open space is the object. Since it is determined whether or not it is a non-existing space, the open space can be recognized with high accuracy.
According to the third aspect of the present invention, there is a region of color information in which the difference from the average value of the color information, which is the spatial color information, is smaller than a preset threshold value, compared with the case where the configuration of the present invention is not provided. Since it is determined that the object is non-existing space, it is determined that the object is a continuous non-existing space even when there is a slight difference in color information due to light adjustment (such as shadows). it can.

According to a fourth aspect of the present invention, an object presence space area in the distance measuring area is specified based on a measurement result by the distance measuring apparatus, and the object is determined based on an image captured by the image capturing apparatus. Spatial color information of the imaging region that overlaps the existence space region is acquired, and it is determined that the region of color information that matches the spatial color information is the object existence space of the imaging region. As a result, the object existence space can be expanded from the object existence space area in the distance measurement area to the imaging area, and the object can be recognized with high accuracy and in a wide range. In addition, the cost of the imaging device, the amount of calculation, and the like can be reduced as compared with the prior art (J01) that performs stereo image analysis using two or more imaging devices.
According to the invention described in claim 5, a commercially available laser range finder can be used as the distance measuring device.

  According to the invention of claim 6, an object non-existing space area in the ranging area is specified based on a measurement result by the ranging apparatus, and based on an image captured by the imaging apparatus, Spatial color information of the imaging region that overlaps the object non-existing space region is acquired, and it is determined that the color information region that matches the spatial color information is the object non-existing space of the imaging region. As a result, the object non-existing space can be expanded from the object non-existing space area in the ranging area to the imaging area, and an open space where no object exists can be recognized accurately and over a wide range. In addition, based on the open space recognized with a wide field of view, the moving path for the moving body to avoid the object can be set more appropriately. In addition, the cost of the imaging device, the amount of calculation, and the like can be reduced as compared with the prior art (J01) that performs stereo image analysis using two or more imaging devices.

  According to the invention of claim 7, an object non-existing space area in the ranging area is specified based on a measurement result by the ranging apparatus, and based on an image captured by the imaging apparatus, Spatial color information of the imaging region that overlaps the object non-existing space region is acquired, and it is determined that the color information region that matches the spatial color information is the object non-existing space of the imaging region. As a result, the object non-existing space can be expanded from the object non-existing space area in the ranging area to the imaging area, and an open space where no object exists can be recognized accurately and over a wide range. it can. In addition, the cost of the imaging device, the amount of calculation, and the like can be reduced as compared with the prior art (J01) that performs stereo image analysis using two or more imaging devices.

  Next, specific examples (examples) of the embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to the following examples.

FIG. 1 is an overall explanatory view of an autonomous mobile body according to a first embodiment of the present invention.
In FIG. 1, the autonomous mobile body AM of the first embodiment includes a mobile body main body M. The movable body M has a front wheel Ma in which the left and right wheels are independently driven, and a rear wheel Mb as an auxiliary wheel of the front wheel Ma. The front wheel Ma is driven by a drive unit M1 built in the movable body main body M. The drive of the drive unit M1 is controlled by a control unit C provided in the movable body main body M.
The front wheel Ma is steered in the left-right direction by controlling the drive of the drive unit M1 by the control unit C so as to give a rotation difference to the two left and right wheels.
Further, the autonomous mobile body AM includes an object such as an obstacle existing in the distance measurement area AR0 by measuring a distance with respect to the distance measurement area AR0 set in front of the autonomous mobile body AM. A laser range finder LRF is supported as an example of a distance measuring device for detecting the distance. The laser range finder LRF measures the distance to the object existing in the distance measuring area AR0 by scanning the laser light in the distance measuring area AR0 and detecting the reflected laser light. In the first embodiment, the distance measurement area AR0 that can be measured by the laser range finder LRF is set in advance so as to be an area having a radius of about 2 m ahead of the autonomous mobile body AM.

In addition, the autonomous moving body AM captures an image including color information with respect to a front imaging area AR1 that overlaps the ranging area AR0 measured by the laser range finder LRF and is wider than the ranging area AR0. A camera CA as an example of an imaging apparatus that captures an image 1 (see FIG. 3A described later) in the area AR1 is supported. In the first embodiment, the camera CA is a commercially available inexpensive web camera. However, the present invention is not limited to this, and a commercially available digital camera, digital video camera, or the like can also be used. .
The environment recognition system S of Example 1 is configured by the control unit C, the laser range finder LRF, and the camera CA.

(Description of the control part C of Example 1)
FIG. 2 is a block diagram (functional block diagram) illustrating each function provided in the control unit of the autonomous mobile body according to the first embodiment of the present invention.
The control unit C is configured by a so-called computer device, and includes a ROM (read only memory) in which programs and data for performing necessary processing are stored, and a RAM (temporarily storing necessary data in RAM ( Random access memory), a CPU (Central Processing Unit) that performs processing according to the program stored in the ROM, and a microcomputer having a clock oscillator and the like, and executes the program stored in the ROM As a result, various functions can be realized.
The control unit C configured as described above can realize various functions by executing programs stored in the ROM. An environment recognition program AP1 is stored in the ROM of the control unit C of the first embodiment.

(Environment recognition program AP1)
The environment recognition program AP1 has the following functional means (program modules).
C1: Distance measurement area distance measurement means The distance measurement area distance measurement means C1 measures the distance to an object such as an obstacle existing in the distance measurement area AR0 by using the laser range finder LRF.

FIG. 3 is an explanatory diagram of an image acquired by the environment recognition program according to the first embodiment of the present invention, FIG. 3A is an explanatory diagram of an image of an imaging region captured by the camera, and FIG. 3B is a measurement result of the laser range finder. FIG. 3C is an explanatory diagram of a state in which the image of the imaging region is compared with the mask image.
C2: Imaging Area Imaging Unit The imaging area imaging unit C2 images the imaging area image 1 (see FIG. 3A) of the imaging area AR1 including color information with the camera CA. The obstacle 1a is imaged in the example of the imaging area image 1 of Example 1 shown in FIG. 3A.

C3: Distance measurement area space specifying means The distance measurement area space specifying means C3 has a mask image generation means C3A, and an object is detected in the distance measurement area AR0 based on the measurement result of the distance measurement area distance measurement means C1. The non-existing object non-existing space area (2a) is specified.
C3A: Mask Image Generation Unit The mask image generation unit C3A is configured to generate a distance measurement area image corresponding to the imaging area image 1 based on the measurement result of the distance measurement area AR0 measured by the distance measurement area distance measurement unit C1. As an example, a mask image 2 (see FIG. 3B) is generated. In FIG. 3B, the mask image generation means C3A of Example 1 is determined from the measurement result of the distance measurement area AR0 by the laser range finder LRF that the obstacle 1a in the imaging area image 1 does not exist. The object non-existing space area 2a is white, and other areas, that is, the object existing space area 2b determined as the obstacle 1a is present, and the non-deniable area 2c that is an area exceeding the distance measuring area AR0, The mask image 2 which is a binary image in which the non-target region 2d (2b + 2c) configured by the above is black is generated.

C4: Spatial Color Information Acquisition Unit The spatial color information acquisition unit C4 includes a mask image collation unit C4A and a color information average value calculation unit C4B, and is based on the imaging region image 1 captured by the imaging region imaging unit C2. Then, the spatial color information, which is the color information of the mask portion 1b (see FIG. 3C) as an example of the overlapping area, which is an area where the imaging area AR1 and the object non-existing space area 2a overlap, is acquired.
C4A: Mask image collating means The mask image collating means C4A collates the imaging area image 1 with the mask image 2 to thereby correspond to the object non-existing space area 2a in the imaging area image 1. The mask part 1b is detected.
C4B: Color information average value calculation means The color information average value calculation means C4B calculates the average value of the color information of the mask portion 1b.
In the spatial color information acquisition unit C4 according to the first exemplary embodiment, the average value of the color information of the mask unit 1b is regarded as the spatial color information.

C5: Spatial discriminating means The spatial discriminating means C5 is based on the imaging area image 1 imaged by the imaging area imaging means C2, and in the imaging area AR1, an area of color information that matches the spatial color information is detected by an object. It is determined that the non-existing object non-existing space 1d (see FIG. 3C). The space discriminating means C5 according to the first embodiment first calculates the average value (spatial color information) of the color information of the mask portion 1b and the non-mask portion 1c (see FIG. 3C) corresponding to the non-target region 2d. When the difference between the color information of the pixels (adjacent regions) adjacent to the mask portion 1b (initial value of the object non-existing space 1d) is smaller than a preset threshold value, the non-mask adjacent to the mask portion 1b By considering the pixels of the part 1c as the object nonexisting space 1d, the object nonexisting space 1d is expanded in the non-masking part 1c.

  When the difference between the expanded color information (space color information) of the object non-existing space 1d and the color information of the pixels of the non-mask portion 1c adjacent to the object non-existing space 1d is smaller than the threshold value, The process of expanding the object nonexisting space 1d is repeated. When the object non-existing space 1d cannot be expanded by all the pixels adjacent to the object non-existing space 1d, the expansion process of the object non-existing space 1d is terminated, and the expanded object non-existing space 1d Discriminated (identified). Note that an example of the object non-existing space 1d according to the first embodiment illustrated in FIG. 3C is the entire area other than the obstacle 1a, more specifically, the floor surface.

C6: Path setting means The path setting means C6 determines that the object non-existing space 1d is a movable area, and sets a moving path that only passes through the object non-existing space 1d and avoids the obstacle 1a. Execute the route setting process.
C7: Movement Control Unit The movement control unit C7 executes a movement control process for moving the movable body main body M along the set movement path. The movement control means C7 of the first embodiment controls the driving unit M1 to control the driving of the front wheel Ma of the movable body main body M, and gives a rotation difference to the left and right wheels of the front wheel Ma. The movement control process is executed by controlling the steering in the left-right direction.

(Description of Flowchart of Example 1)
Next, the processing flow of the environment recognition program AP1 of the control unit C according to the first embodiment will be described with reference to a flowchart. As for the movement control process corresponding to the movement control means C7, the route setting process (FIG. 4 to be described later) is controlled by controlling the driving and steering (adjustment of the movement direction) of the front wheels Ma of the movable body M. Since only the moving body M is moved along the movement route set in (see the flowchart), a detailed description of the flowchart is omitted.
(Description of Flowchart of Route Setting Process of Example 1)
FIG. 4 is a flowchart of the path setting process of the environment recognition program according to the first embodiment of the present invention.
The processing of each ST (step) in the flowchart of FIG. 4 is performed according to a program stored in the ROM or the like of the control unit C. Further, this process is executed by multitasking in parallel with other various processes (for example, movement control process) of the control unit C.

The flowchart shown in FIG. 4 is started when the environment recognition program AP1 is started after the autonomous mobile body AM is powered on.
In ST1 of FIG. 4, the following processes (1) and (2) are executed, and the process proceeds to ST2.
(1) The distance measurement area AR0 is measured by the laser range finder LRF.
(2) The imaging area image 1 (see FIG. 3A) of the imaging area AR1 is captured by the camera CA.
In ST2, a mask image 2 (see FIG. 3B) is generated based on the measurement result of the ranging area AR0 by the laser range finder LRF and the imaging area image 1 of the imaging area AR1 by the camera CA. Then, the process proceeds to ST3.
In ST3, the imaging area image 1 and the mask image 2 are collated to detect the mask portion 1b (see FIG. 3C) of the imaging area image 1, and the average value of the color information of the mask portion 1b is calculated. Then, the process proceeds to ST4.

In ST4, it is determined whether or not the difference between the average value (spatial color information) of the color information of the mask portion 1b and the color information of the pixels of the adjacent non-mask portion 1c is smaller than a preset threshold value. If yes (Y), the process proceeds to ST5, and, if no (N), the process proceeds to ST6.
In ST5, the object non-existing space 1c is expanded in the non-masking part 1c by setting the pixels of the adjacent non-masking part 1c as the object non-existing space 1d (see FIG. 3C). Then, the process proceeds to ST6.
In ST6, it is determined whether or not it has been confirmed that the object nonexistence space 1d cannot be expanded by all the pixels adjacent to the object nonexistence space 1d. If yes (Y), the process proceeds to ST7, and, if no (N), the process returns to ST4.
In ST7, it is determined that the object non-existing space 1d is a movable region, and a moving path that passes only the object non-existing space 1d and avoids the obstacle 1a is set. Then, the process returns to ST1.

(Operation of Example 1)
In the autonomous mobile AM according to the first embodiment having the above-described configuration, in the imaging region AR1, a region of color information that matches the spatial color information that is an average value of the color information of the mask unit 1b is the object non-display. The existence space 1d is determined (see ST4 to ST6 in FIGS. 3C and 4). Therefore, by combining the information of the laser range finder LRF of the narrow ranging area AR0 and the information of the camera CA of the wide imaging area AR1, the object non-existing space 1d of the ranging area AR0 measured by the highly accurate laser range finder LRF. Can be expanded into the imaging area AR1.

Further, in the autonomous mobile object AM according to the first embodiment, the spatial color information of the object non-existing space 1d (the average value of the color information of the mask unit 1b) and the non-masking unit 1c adjacent to the object non-existing space 1d. When the difference from the color information of the pixel is smaller than the threshold value, the process of expanding the object non-existing space 1d is performed until the object non-existing space 1d cannot be expanded in all pixels adjacent to the object non-existing space 1d. Repeat (see ST4 to ST6 in FIG. 4). Therefore, in the autonomous mobile AM according to the first embodiment, since it is not determined whether or not the region that is not adjacent to the object nonexisting space 1d is the object nonexisting space 1d, the object nonexisting space 1d is determined. For example, whether or not the object non-existence space is only for the area adjacent to the open space without determining the closed space such as the object non-existence space surrounded by the object. Therefore, it is possible to accurately recognize the open space, that is, the object non-existing space 1d. Further, since it is determined that the area of color information whose difference from the spatial color information (average value of the color information of the mask portion 1b) is smaller than a preset threshold is the object non-existing space 1d, for example, the same Even in the case where an object is slightly different in color information due to light adjustment (such as object shadow), it can be determined that the object is a continuous non-existing space.
Further, in the autonomous mobile body AM according to the first embodiment, it is determined that the object non-existing space 1d is a movable area, and only the object non-existing space 1d is passed to avoid the obstacle 1a. Can be set (see ST7 in FIG. 4).

(Experimental example)
Here, it is examined how the accuracy of recognizing the surrounding environment is changed by the autonomous mobile body AM according to the first embodiment executing the route setting process (see ST1 to ST7 in FIG. 4). Therefore, the following Experimental Example 1 and Comparative Example 1 were prepared.
(Experimental example 1)
The autonomous mobile AM of Experimental Example 1 is a simulator manufactured with the same configuration as the autonomous mobile AM of Example 1, and the route setting process and the movement control process of Example 1 are executed. . Therefore, in the autonomous mobile AM of Experimental Example 1, the object non-existing space 1d in the imaging region image 1 is expanded from the mask portion 1b corresponding to the object non-existing space region 2a to the non-mask portion 1c. It is determined (see FIG. 3C). That is, the movable area is determined by extending from the object non-existing space area 2a (see FIG. 3B) of the distance measuring area AR0 specified by the laser range finder LRF to the imaging area AR1.

(Comparative Example 1)
The autonomous mobile body AM ′ of the comparative example 1 does not have the camera CA as compared with the autonomous mobile body AM of the experimental example 1, and the route setting process of the comparative example 1 is the route setting process of the experimental example 1. In contrast, the object non-existing space 1d in the imaging region image 1 is determined to be only the mask portion 1b corresponding to the object non-existing space region 2a, and is expanded to the non-mask portion 1c and is not determined. That is, in the autonomous moving body AM ′ of the comparative example 1, it is determined that the movable area is only the object non-existing space area 2a of the distance measuring area AR0 specified by the laser range finder LRF.

(Experimental conditions)
FIG. 5 is an explanatory diagram of experimental conditions and experimental results of the experimental example, FIG. 5A is a diagram illustrating experimental conditions of the experimental example, and FIG. 5B is a diagram illustrating experimental results of the experimental example. It is the figure which showed the path | route which the type | mold mobile body traveled by the continuous line, and showed the path | route which the autonomous mobile body of the comparative example 1 drive | worked with the broken line.
As shown in FIG. 5A, in the experimental example, a simulation is performed in which the traveling course 3 of the autonomous mobile bodies AM and AM ′ of Experimental Example 1 and Comparative Example 1 is provided and the traveled route and the travel time for one lap are measured. The experiment was conducted.
Here, the round course 3 is a boundary between the floor surface 3a on which the autonomous mobile bodies AM and AM ′ travel and the floor surface 3a, and an outer curb 3b and an inner curb surrounding the floor 3a. 3c and an obstacle 3d provided on the outer curb 3b side. The floor surface 3a is set in advance so that color information is different from that of the curbs 3b and 3c and the obstacle 3d.

(Experimental result)
As shown in FIG. 5B, the route traveled by the autonomous mobile body AM in Experimental Example 1 (see the solid line in FIG. 5B) is the route traveled by the autonomous mobile body AM ′ in Comparative Example 1 (see the broken line in FIG. 5B). It can be seen that the distance from the outer curb 3b and the obstacle 3d is shorter than that of the outer curb 3b. In addition, the traveling time of the autonomous mobile body AM of Experimental Example 1 was 26.38 seconds, and the traveling time of the autonomous mobile body AM ′ of Comparative Example 1 was 33.26 seconds. That is, it can be seen that the traveling time of the autonomous mobile body AM of Experimental Example 1 is improved by about 26% compared to the traveling time of the autonomous mobile body AM ′ of Comparative Example 1.

Therefore, the autonomous mobile body AM according to the first embodiment can complement the narrowness of the distance measurement area AR0 by the laser range finder LRF by the width of the imaging area AR1 by the camera CA. The low accuracy of recognizing an object can be complemented by the high accuracy of recognizing the object by the laser range finder LRF. Therefore, it is possible to recognize an accurate and wide range of the obstacle and the movable space, and it is possible to widen the field of view of the autonomous mobile body AM. Therefore, the movement route can be set more appropriately based on a wide field of view.
As a result, the autonomous mobile body AM according to the first embodiment has the object non-existing space 1d as compared to the case where the information of the laser range finder LRF is not supplemented by the camera CA like the autonomous mobile body AM ′ according to the first comparative example. Therefore, the surrounding environment can be recognized with high accuracy, and the movement route can be set more appropriately.

  Furthermore, since the camera CA of the first embodiment only captures the imaging area image 1 including color information, it is possible to use a commercially available inexpensive Web camera. Therefore, the autonomous mobile body AM according to the first embodiment is an obstacle 1a in front of the autonomous mobile body AM when the prior art (J01) is applied, that is, stereo images are analyzed using two expensive cameras. Compared to the case where the object non-existing space 1c is detected based on the calculation result obtained by calculating the distance up to, the cost of the camera CA and the calculation amount for recognizing the surrounding environment can be reduced.

(Example of change)
As mentioned above, although the Example of this invention was explained in full detail, this invention is not limited to the said Example, A various change is performed within the range of the summary of this invention described in the claim. It is possible. Modification examples (H01) to (H04) of the present invention are exemplified below.
(H01) In an embodiment of the present invention, the route setting process (see ST1 to ST7 in FIG. 4) specifies the object non-existing space area 2a in the ranging area AR0 by the laser range finder LRF, and the object The non-existing space 1d is expanded from the ranging area AR0 to the imaging area AR1, and it is determined that the object non-existing space 1d is a movable area, but the present invention is not limited to this, and the laser range finder LRF Is used to identify the object existence space area 2b in the ranging area AR0 and determine that the color information area that matches the color information (spatial color information) of the object existence space area 2b is the object existence space where the object exists. It is also possible to set a movement route that avoids the object existence space (obstacle 1a). Further, when determining the object existence space instead of the object non-existence space 1d, the object existence space is not limited to a configuration in which a movement path that avoids the object existence space as the obstacle 1a is set. For example, the object as a moving object It is also possible to adopt a configuration in which a movement route that follows the existence space is set.

(H02) In the embodiment of the present invention, the laser range finder LRF is used as a distance measuring device that detects an object existing in the distance measuring area by measuring a distance with respect to a predetermined distance measuring area. However, the present invention is not limited to this, and other devices such as a scanning laser radar and a millimeter wave radar (see Patent Document 1) can also be used.
(H03) In an embodiment of the present invention, the color information (spatial color information) of the object non-existing space 1d is compared with the color information of the pixels of the non-mask portion 1c adjacent to the object non-existing space 1d. Although the process of expanding the object non-existing space 1d is repeated (see ST4 to ST6 in FIG. 4), the present invention is not limited to this. For example, from the pixel of the non-mask part 1c that is not adjacent to the object non-existing space 1d In comparison, the object non-existing space 1d can be expanded.
(H04) In the embodiment of the present invention, the mobile body M having the environment recognition system S is configured by a so-called robot car, but is not limited thereto, and is configured by, for example, a bipedal walking robot or the like. It is also possible.

FIG. 1 is an overall explanatory view of an autonomous mobile body according to a first embodiment of the present invention. FIG. 2 is a block diagram (functional block diagram) illustrating each function provided in the control unit of the autonomous mobile body according to the first embodiment of the present invention. FIG. 3 is an explanatory diagram of an image acquired by the environment recognition program according to the first embodiment of the present invention, FIG. 3A is an explanatory diagram of an image of an imaging region captured by the camera, and FIG. 3B is a measurement result of the laser range finder. FIG. 3C is an explanatory diagram of a state in which the image of the imaging region is compared with the mask image. FIG. 4 is a flowchart of the path setting process of the environment recognition program according to the first embodiment of the present invention. FIG. 5 is an explanatory diagram of experimental conditions and experimental results of the experimental example, FIG. 5A is a diagram illustrating experimental conditions of the experimental example, and FIG. 5B is a diagram illustrating experimental results of the experimental example. It is the figure which showed the path | route which the type | mold mobile body traveled by the continuous line, and showed the path | route which the autonomous mobile body of the comparative example 1 drive | worked with the broken line.

Explanation of symbols

1 ... Image,
1a ... object,
1d: object non-existence space,
2a ... object non-existence space area,
2b ... object existence space region,
AM ... Autonomous moving body,
AP1 ... environmental recognition program,
AR0: ranging area,
AR1 ... imaging area,
C ... Computer,
C3: Ranging area space specifying means,
C4: Spatial color information acquisition means,
C5: Spatial discrimination means,
C6: route setting means,
C7: Movement control means,
CA: imaging device,
LRF: Ranging device, laser range finder,
M ... mobile,
S: Environment recognition system.

Claims (7)

A distance measuring device for detecting an object existing in the distance measuring area by measuring a distance with respect to a predetermined distance measuring area;
An imaging device that captures an image including color information with respect to an imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object non-existing space area in which no object exists in the ranging area based on a measurement result by the ranging device;
Based on the image captured by the imaging device, spatial color information acquisition means for acquiring spatial color information that is color information of the object non-existing spatial region and an overlapping imaging region;
Space discriminating means for discriminating an area of color information matching the spatial color information in the imaging area as an object non-existing space based on an image captured by the imaging apparatus;
An environment recognition system characterized by comprising: Based on the image captured by the imaging device, when the spatial color information matches the color information of the area adjacent to the area of the spatial color information in the imaging area, The space discriminating means for discriminating that the space exists;
The environment recognition system according to claim 1, further comprising: Based on the image captured by the imaging device, the spatial color information acquisition unit that acquires, as the spatial color information, an average value of the color information of the object non-existing space region and the overlapping imaging region;
Based on the image captured by the imaging device, the color information matches the spatial color information when there is color information in the imaging region that is less than a preset threshold value with respect to the spatial color information. The space discriminating means which regards the area of the color information as the color information and discriminates the area of the color information as the object non-existing space;
The environment recognition system according to claim 1 or 2, further comprising: A distance measuring device for detecting an object existing in the distance measuring area by measuring a distance with respect to a predetermined distance measuring area;
An imaging device that captures an image including color information with respect to an imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object existence space area where an object exists in the ranging area based on a measurement result by the ranging device;
Spatial color information acquisition means for acquiring spatial color information that is color information of the imaging region that overlaps the object presence space region based on an image captured by the imaging device;
Space discriminating means for discriminating, in the imaging area, an area of color information that matches the spatial color information as an object existence space based on an image captured by the imaging apparatus;
An environment recognition system characterized by comprising: The distance measuring device configured by a laser range finder that measures a distance to an object existing in the distance measuring area by scanning laser light in the distance measuring area and detecting laser light reflected by the object. When,
The environment recognition system according to any one of claims 1 to 4, further comprising: A distance measuring device that detects an object existing in the distance measuring area by measuring a distance with respect to a distance measuring area set in front of the moving body;
An imaging device that captures an image including color information with respect to a front imaging region that overlaps with a ranging region to be measured by the ranging device and is wider than the ranging region;
Ranging area space specifying means for specifying an object non-existing space area in which no object exists in the ranging area based on a measurement result by the ranging device;
Based on the image captured by the imaging device, spatial color information acquisition means for acquiring spatial color information that is color information of the object non-existing spatial region and an overlapping imaging region;
Space discriminating means for discriminating an area of color information matching the spatial color information in the imaging area as an object non-existing space based on an image captured by the imaging apparatus;
It is determined that the object non-existing space is a movable area, and a route setting means for setting a moving route that avoids the object;
Movement control means for moving the moving body along the set movement path;
An autonomous mobile body characterized by comprising: An object in which no object exists in the distance measurement area based on a measurement result by a distance measuring device that detects an object existing in the distance measurement area by measuring a distance with respect to a predetermined distance measurement area. Ranging area space specifying means for specifying a non-existing space area,
The object non-existing space is based on an image captured by an imaging device that captures an image including color information for an imaging region that overlaps a ranging region measured by the ranging device and is wider than the ranging region. Spatial color information acquisition means for acquiring spatial color information which is color information of an area and an overlapping imaging area;
Space discriminating means for discriminating, based on an image captured by the imaging device, an area of color information that matches the spatial color information in the imaging area as an object non-existing space in which no object exists;
Environment recognition program to function as