JP7063760B2 - Mobile - Google Patents

Description

The present invention relates to a moving body that acquires a current position using information acquired by a sensor.

Conventionally, running control based on a comparison result between a plurality of teaching images and an actual image acquired by a photographing device and running control based on a detection result of a distance and a direction to a surrounding object are set as predetermined conditions. A moving body that switches according to the situation is known (see Patent Document 1).

In Patent Document 1, for example, the number of feature points that have succeeded in pattern matching between an captured image and an imaging landmark is equal to or less than a predetermined threshold, or the number of recognized imaging landmarks is a plurality. , If it is not possible to determine which landmark measurement result should be used, or if the amount of deviation from the angle from the traveling direction obtained from the captured image exceeds a predetermined range, driving control based on the comparison result of the images. Therefore, it is exemplified that the vehicle can be switched to the travel control based on the detection result such as the distance. Therefore, in the conventional example described in Patent Document 1, when a predetermined condition regarding one of the two running controls is satisfied, the other running control is switched to.

Japanese Unexamined Patent Publication No. 2010-140246

However, in the conventional moving body, when a predetermined condition for one running control is satisfied, even if the other running control is switched, it is not always possible to perform appropriate running control after the switching. was there. For example, when the number of feature points that have succeeded in pattern matching between the captured image and the landmark for imaging is less than or equal to a predetermined threshold value, even if the driving control is switched to based on the detection result such as the distance, the surroundings are around. When the moving object is located in a place where the object to be measured does not exist, it becomes impossible to perform appropriate traveling control based on the detection result such as the distance.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a moving body capable of appropriately acquiring a current position by using a plurality of position estimates.

In order to achieve the above object, the moving body according to the present invention is a moving body that moves autonomously, and is a first sensor that acquires information around the moving body and information about the surroundings of the moving body. The second sensor that acquires information different from the information acquired by the first sensor, and the first estimated position by estimating the position of the moving object using the information acquired by the first sensor. The second position is estimated by using the information acquired by the second sensor and the first position estimation unit that acquires the first reliability corresponding to the first estimated position. The second position estimation unit that acquires the estimated position and the second reliability corresponding to the second estimated position, the first and second estimated positions, and the first and second reliability. It is provided with a current position acquisition unit for acquiring the current position of the moving body, a moving mechanism for moving the moving body, and a movement control unit for controlling the moving mechanism using the current position.
With such a configuration, since the current position of the moving body can be acquired using the first and second reliability, for example, the current position can be acquired by comparing both reliabilitys. As a result, for example, in a situation where the first reliability is not high but the second reliability is even lower than the first reliability, the current position is acquired using the first estimated position. Will be able to.

Further, in the moving body according to the present invention, the first sensor is a distance measuring sensor that measures the distance to a surrounding object in a plurality of directions, and the first position estimation unit is the distance measured by the first sensor. , And the first estimated position is acquired using the first map showing the position of the obstacle in the moving area of the moving body, and the second sensor is an image sensor that acquires a photographed image around the moving body. , The second position estimation unit uses a second map showing the positions of the feature points identified in the captured image acquired by the second sensor and the feature points in the moving region of the moving body, and the second estimated position. May be obtained.
With such a configuration, it is possible to perform position estimation using the distance measurement result and position estimation using the imaging result, and it becomes possible to acquire the current position using those estimation results. Normally, the position estimation using the distance measurement result has a feature that the accuracy is low when the distance to the surrounding object is long, whereas the change in the ambient brightness is robust. On the other hand, the position estimation using the shooting result is characterized in that the position estimation can be performed with high accuracy even when the distance to the surrounding object is long, but it is vulnerable to the change in the ambient brightness. As described above, since both are complementary, performing position estimation using both is effective for obtaining the current position with higher accuracy.

Further, in the moving body according to the present invention, the second position estimation unit acquires the appropriateness indicating the degree of appropriateness of the feature points for each feature point, and is divided into a plurality of blocks using the appropriateness. The block reliability is calculated for each block of the captured image, the second estimated position is acquired using the block with high block reliability, and the block reliability of the block used in the acquisition of the second estimated position is used to obtain the second estimated position. You may acquire the reliability of 2.
With such a configuration, it is possible to perform position estimation using an appropriate block portion of the captured image, and it is possible to realize highly accurate position estimation using the captured image. As described above, the position estimation using the shooting result is usually vulnerable to the change in brightness, but even if the brightness changes, it does not mean that the entire area of the shot image cannot be used for the position estimation. Therefore, by doing so, it is effective to perform position estimation using the feature points of the block satisfying a predetermined condition among the feature points included in the captured image.

Further, in the moving body according to the present invention, the current position acquisition unit may set the estimated position corresponding to the higher reliability of the first and second reliability as the current position of the moving body.
With such a configuration, it is possible to avoid using the result of the position estimation of the other automatically when the position estimation of one is inappropriate as in the above-mentioned conventional example, and the accuracy is higher. It is possible to acquire the current position.

Further, in the moving body according to the present invention, the current position acquisition unit may synthesize the current position of the moving body from the first and second estimated positions by using the first and second reliability.
With such a configuration, for example, when the first reliability is high, the current position close to the first estimated position is acquired, or when the first and second reliability are the same, the first reliability is the same. It is possible to obtain the current position between the first and second estimated positions.

According to the moving body according to the present invention, the current position can be appropriately acquired by using a plurality of position estimates.

A block diagram showing a configuration of a moving body according to an embodiment of the present invention. A flowchart showing the operation of the moving body according to the same embodiment. Top view showing an example of the distance measuring sensor included in the moving body in the same embodiment. Top view showing an example of an image sensor of a moving body in the same embodiment. Top view showing an example of two ranging sensors included in a moving body in the same embodiment. Top view showing an example of two image sensors possessed by a moving body in the same embodiment. Schematic diagram showing an example of distance measurement in the same embodiment Schematic diagram showing an example of a block of captured images in the same embodiment

Hereinafter, the moving body according to the present invention will be described with reference to embodiments. In the following embodiments, the components and steps with the same reference numerals are the same or correspond to each other, and the description thereof may be omitted again. The moving body according to the present embodiment acquires the current position using the first and second estimated positions acquired by using two sensors and the first and second reliability corresponding to them, respectively. It is something to do.

FIG. 1 is a block diagram showing a configuration of a moving body 1 according to the present embodiment. The moving body 1 according to the present embodiment moves autonomously, and includes a first sensor 11, a second sensor 12, a first position estimation unit 13, and a second position estimation unit 14. A current position acquisition unit 15, a movement mechanism 16, and a movement control unit 17 are provided. Note that the autonomous movement of the moving body 1 may mean that the moving body 1 does not move according to an operation instruction received from a user or the like, but moves to a destination at its own discretion. The destination may be, for example, manually determined or automatically determined. Further, the movement to the destination may or may not be performed along the movement route, for example. Further, moving to the destination by one's own judgment may mean, for example, moving to the destination by the moving body 1's own judgment of the traveling direction, movement, stop, and the like. Further, for example, the moving body 1 may move so as not to collide with an obstacle. The moving body 1 may be, for example, a dolly or a moving robot. The robot may be, for example, an entertainment robot, a monitoring robot, a transfer robot, a cleaning robot, or a robot that shoots a moving image or a still image. , Other robots may be used.

The first and second sensors 11 and 12 acquire information around the moving body 1. The first and second sensors 11 and 12 may be distance measuring sensors that independently measure the distance to a surrounding object in a plurality of directions, and are images for acquiring a captured image of the surroundings of the moving body 1. It may be a sensor. When the first and second sensors 11 and 12 are distance measuring sensors, the information to be acquired is a distance, and when the first and second sensors 11 and 12 are image sensors, the information to be acquired is an image sensor. The information of is a photographed image. The second sensor 12 acquires information different from the information acquired by the first sensor 11. The information acquired by the first and second sensors 11 and 12 may be different, for example, (1) the types of the first and second sensors 11 and 12 may be different, and (2) first. The information acquisition directions of the first and second sensors 11 and 12 may be different. Hereinafter, each case will be described.

(1) When the types of the first and second sensors 11 and 12 are different In this case, for example, the first sensor 11 may be a distance measuring sensor and the second sensor 12 may be an image sensor. .. 3A and 3B are top views of the moving body 1, and are views showing distance measurement by the first sensor 11 which is a distance measurement sensor and photographing by a second sensor 12 which is an image sensor, respectively. In FIGS. 3A and 3B, the distance measuring sensor and the image sensor are shown in separate drawings for convenience of explanation, but in reality, the distance measuring sensor and the image sensor are shown in one moving body 1. It will be installed. The direction of distance measurement of the first sensor 11 which is a distance measurement sensor and the direction of photography of the second sensor 12 which is an image sensor are the same, for example, as shown in FIGS. 3A and 3B. May or may be different. When the distance measuring direction of the first sensor 11 and the photographing direction of the second sensor 12 are the same, the direction may be the traveling direction of the moving body 1. This is because it is possible to perform control such as collision avoidance by using the distance measurement result in the traveling direction, and the captured image in the traveling direction has a smaller change, so that it is suitable for use in position estimation.

(2) When the information acquisition directions of the first and second sensors 11 and 12 are different In this case, for example, both the first and second sensors 11 and 12 may be distance measuring sensors. Alternatively, it may be an image sensor. FIG. 3C is a top view showing distance measurement by the first and second sensors 11 and 12, which are distance measurement sensors. FIG. 3D is a top view showing photography by the first and second sensors 11 and 12, which are image sensors. As shown in FIGS. 3C and 3D, when both the first and second sensors 11 and 12 are sensors of the same type, it is assumed that the information acquisition directions are different. It should be noted that the fact that the information acquisition directions are different may be considered to mean that the information acquired by the first and second sensors 11 and 12 includes at least information in different directions. .. That is, at least a part of the information acquired by the first and second sensors 11 and 12 may be in the same direction. Note that FIGS. 3C and 3D show a case where the information acquisition directions of the first and second sensors 11 and 12 differ by 180 degrees, but this may not be the case. The information acquisition directions of the first and second sensors 11 and 12 may differ by, for example, 90 degrees.

The first and second sensors 11 and 12 may be of different types and may have different information acquisition directions. Further, in FIGS. 3A to 3D, the case where the contour shape of the moving body 1 in the top view is trapezoidal is shown, but this is to make the orientation of the moving body 1 easy to understand, and the moving body 1 has a trapezoidal shape. The contour shape in the top view may be, for example, a rectangular shape, a circular shape, or a polygonal shape, and is not particularly limited. In the present embodiment, the case where the first sensor 11 is a distance measuring sensor and the second sensor 12 is an image sensor will be mainly described.

The distance measuring sensor may be, for example, a laser sensor, an ultrasonic sensor, a distance sensor using a microwave, a distance sensor using a stereo image taken by a stereo camera, or the like. The laser sensor may be a laser range sensor (laser range scanner). It should be noted that these distance measuring sensors are already known, and their description will be omitted. In this embodiment, the case where the ranging sensor is a laser range sensor will be mainly described. Further, the first sensor 11 may have one laser range sensor, or may have two or more laser range sensors. In the latter case, two or more laser range sensors may cover all directions. Further, when the distance measuring sensor is an ultrasonic sensor, a distance sensor using a microwave, or the like, the distance in a plurality of directions may be measured by rotating the distance measuring direction of the distance measuring sensor, and the distance may be measured in each of a plurality of directions. A plurality of distance measuring sensors arranged in may be used to measure distances in a plurality of directions. Measuring the distance in a plurality of directions may mean, for example, measuring the distance in a plurality of directions at a predetermined angle interval for a predetermined angle range or the entire circumference (360 degrees). The angle interval may be constant, for example, 1 degree interval, 2 degree interval, 5 degree interval, and the like. The information obtained from the distance measuring sensor may be, for example, the distance to a surrounding object with respect to each of a plurality of azimuth angles with respect to a certain direction of the moving body 1. By using the distance, it becomes possible to know what kind of object exists around the moving body 1 in the local coordinate system of the moving body 1.

The image sensor may be, for example, a CCD image sensor, a CMOS image sensor, or the like. The image sensor may include, for example, an optical system such as a lens for forming an image on the image sensor. Further, the image sensor may be monocular or binocular (stereo camera). In this embodiment, the case where the image sensor is monocular will be mainly described. The image sensor usually captures a moving image, that is, acquires a continuous image frame. Since the captured image is acquired in order to acquire the current position of the moving moving body 1, it is preferable that the frame rate is sufficiently large with respect to the moving speed. For example, the frame rate may be about 30 fps or the like.

It is preferable that the acquisition of information by the first and second sensors 11 and 12 is performed synchronously. This is because it is preferable that the positions of the simultaneous points are estimated by the first and second position estimation units 13 and 14, respectively.

The first position estimation unit 13 estimates the position of the moving body 1 by using the information acquired by the first sensor 11, and thereby corresponds to the first estimated position and the first estimated position. Obtain the reliability of 1. As described above, in the present embodiment, in order to explain the case where the first sensor 11 is a distance measuring sensor, the first position estimation unit 13 uses the distance measured by the first sensor 11 and the distance measured by the first sensor 11. It is assumed that the first estimated position is acquired by using the first map showing the position of the obstacle in the moving area of the moving body 1. Strictly speaking, the measured distance is the measured distance in each direction. Further, it is assumed that the first map is stored in a recording medium (not shown) included in the first position estimation unit 13. The obstacle may be an object such as a wall or equipment for which the distance is measured by the distance measuring sensor. The first estimated position is usually a position in the world coordinate system. Since self-position estimation using such distance measurement results is already known in SLAM (Simultaneous Localization and Mapping) using distance measurement results, detailed description thereof will be omitted. The first position estimation unit 13 may or may not create the first map, which is an environmental map. In the former case, the first position estimation unit 13 may be the same as the SLAM using the distance measurement result, and in the latter case, the first position estimation unit 13 may be the existing environment. The self-position estimation may be performed using the first map which is a map.

The first reliability indicates the reliability of the first estimated position, that is, the certainty, and may be acquired according to the degree of matching between the measurement result and the first map, for example, and is measured. It may be obtained depending on the accuracy of the result, it may be both, or it may be obtained by other methods. Therefore, the first reliability is, for example, a matching between the measurement result by the first sensor 11 and the distance to the obstacle shown by the first map when the moving body 1 is present at the first estimated position. The higher the degree of, the larger the value may be. In this case, the first reliability may be calculated as follows.
First reliability = ( _ΣF (Li)) / N1

Here, the function F (x) is a decreasing function that becomes 1 when x = 0 and becomes smaller as x becomes a positive larger value. However, F (x) ≧ 0. L _i is the absolute value of the difference between the distance measurement result in the i-th distance measurement direction and the distance to the edge of the obstacle shown by the first map. Further, N1 is the total number of i, that is, the total number of distance measurements, and the sum Σ is taken for all i. Since it is standardized by N1, the first reliability is a real number of 0 or more and 1 or less. Note that F (x) may be any function, and may be, for example, F (x) = exp (−α × x). α is a positive real number.

Further, since the error of the distance measurement result by the laser or the like becomes larger due to the influence of reflection or the like as the distance becomes longer, the first reliability may be, for example, a smaller value as the distance measurement result is larger. .. In this case, the first reliability may be calculated as follows.
First reliability ₌ (ΣG (Di)) / N1

Here, the function G (x) is a decreasing function that becomes 1 when x = 0 and becomes smaller as x becomes a positive larger value. However, G (x) ≧ 0. Further, D _i is a distance measurement result in the i-th distance measurement direction. Further, N1 and the total sum Σ are as described above. Also in this case, since it is standardized by N1, the first reliability is a real number of 0 or more and 1 or less. Further, G (x) may be the same as the above function F (x). Further, the first reliability is, for example, the higher the degree of matching between the measurement result by the first sensor 11 and the position of the obstacle shown in the first map, the larger the value, and the larger the distance measurement result. The value may be as small as possible.

The first position estimation unit 13 may set the first reliability to a low value even when the first estimated position cannot be uniquely determined. For example, when the moving body 1 exists in a monotonous environment such as a long corridor, it may not be possible to uniquely determine the first estimated position by using the distance measurement result by the first sensor 11. In such a case, the first reliability may be set to a low value.

The second position estimation unit 14 estimates the position of the moving body 1 by using the information acquired by the second sensor 12, so that the second estimated position and the second estimated position correspond to the second estimated position. Get the reliability of 2. As described above, in order to explain the case where the second sensor 12 is an image sensor in the present embodiment, the second position estimation unit 14 is specified in the captured image acquired by the second sensor 12. It is assumed that the second estimated position is acquired by using the second map showing the position of the feature point and the feature point in the moving region of the moving body 1. It is assumed that the second map is stored in a recording medium (not shown) included in the second position estimation unit 14. The second estimated position is usually a position in the world coordinate system. Since self-position estimation using such captured images is already known as visual-SLAM, a detailed description thereof will be omitted. The second position estimation unit 14 may or may not create the second map, which is an environmental map. In the former case, the second position estimation unit 14 may be similar to visual-SLAM, and in the latter case, the second position estimation unit 14 is an existing environmental map. The self-position estimation may be performed using the map of 2.

Here, the feature points specified by the second position estimation unit 14 in the captured image will be described. When the second position estimation unit 14 performs self-position estimation similar to ORB-SLAM, a predetermined number of FAST key points are used as feature points. The predetermined number of FAST key points is selected from the FAST key points acquired from the captured image, which is higher than the Harris corner measure. For the FAST key points, Harris's corner scale, and feature points in ORB-SLAM, refer to the following documents 1 to 3, respectively. Further, as the feature point, for example, a SIFT key point, a SURF key point, or the like may be used. For those key points, refer to the following documents 4 and 5. Moreover, other feature points used in visual-SLAM may be used. Further, for the feature descriptor of the feature point described later, for example, refer to the following documents 3 to 5. In the present embodiment, a case where the second position estimation unit 14 performs self-position estimation by the same method as ORB-SLAM will be mainly described. For ORB-SLAM, refer to Document 6 below.

Reference 1: Edward Rosten, Tom Drummond, "Machine learning for high-speed corner detection", in 9th European Conference on Computer Vision, vol. 1, 2006, pp. 430-443.
Reference 2: C. Harris, M. Stephens, "A Combined Corner and Edge Detector", In Proceedings of the 4th Alvey Vision Conference (1988), pp. 147-151.
Reference 3: Ethan Rublee, Vincent Rabaud, Kurt Konolige, Gary R. Bradski, "ORB: An efficient alternative to SIFT or SURF", ICCV 2011, pp.2564-2571.
Reference 4: D. Lowe, "Distinctive image features from scale-invariant keypoints", Int. Journal of Computer Vision (2004), 60 (2), pp.91-110.
Reference 5: H. Bay, A. Ess, T. Tuytelaars, L. Van Gool, "SURF: Speeded Up Robust Features", Computer Vision and Image Understanding (CVIU) (2008), 110 (3), pp.346- 359. 359.
Reference 6: R. Mur-Artal, JM Montiel, JD Tardos, "ORB-SLAM: A Versatile and Accurate Monocular SLAM System", IEEE Transactions on Robotics 31 (2015), pp.1147-1163.

The second reliability indicates the reliability of the second estimated position, that is, the certainty, and is, for example, the matching between the feature points specified in the captured image and the feature points included in the second map. It may be acquired according to the degree, may be acquired according to the accuracy of the feature points specified in the captured image, may be both, or may be acquired by other methods. Therefore, the second reliability is, for example, that the feature descriptor (feature amount) of the feature point specified in the captured image acquired by the second sensor 12 and the moving body 1 are present at the second estimated position. In this case, the larger the similarity with the feature descriptor of the feature points included in the second map corresponding to those feature points, the larger the value may be. In that case, the second reliability may be calculated as follows.
Second reliability = (ΣR _i ) / N2

Here, R _i is the degree of similarity between the feature descriptor of the i-th feature point specified in the captured image and the feature descriptor of the feature point included in the second map corresponding to the feature point. It is preferable that the value is standardized from 0 to 1. As for the degree of similarity, the larger the value, the greater the degree of similarity. Further, when the feature point corresponding to the i-th feature point specified in the captured image is not included in the second map, the similarity may be 0. Further, N2 is the total number of i, that is, the total number of feature points specified in the captured image, and the sum Σ is taken for all i. Since it is standardized by N2, the second reliability is a real number of 0 or more and 1 or less. When the feature descriptor is a binary vector (for example, in the case of ORB-SLAM), the Hamming distance may be used to calculate the similarity. If the feature descriptor is not a binary vector (eg, SIFT keypoints or SURF keypoints), the similarity may be calculated using the cosine distance or the Euclidean distance.

In addition, the feature points in the captured image cannot be properly identified when the exposure is overexposed and the image is overexposed, or when the exposure is underexposed and the image is blacked out. The degree may be, for example, the larger the evaluation value used in specifying the feature points, the larger the value. It is assumed that a candidate for a feature point having a high evaluation value is specified as a feature point. For example, in ORB-SLAM, since a predetermined number of FAST key points are selected from the upper ranks of Harris's corner scale, the evaluation value may be Harris's corner scale. That is, the second reliability may be a representative value of Harris's corner scale with respect to the plurality of feature points specified in the captured image. The representative value may be, for example, an average value, a median value, a maximum value, or the like. Further, for example, in the identification of feature points (key points) by SIFT, in order to delete low-contrast key point candidates, the size of the extreme value of the key point candidate points is calculated, and the candidate points whose size is less than the threshold value are calculated. To delete. Therefore, a representative value of the magnitude of the extremum with respect to the specified plurality of feature points may be used as the second reliability. The same applies to other feature points. Further, the second reliability is, for example, the similarity between the feature descriptor of the feature points specified in the captured image and the feature descriptor of the feature points included in the second map corresponding to those feature points. The larger the value, the larger the value, and the larger the evaluation value used in specifying the feature point, the larger the value.

It is preferable that the first and second reliability are standardized so that the two can be appropriately compared. That is, if the first and second reliability values are the same, it is preferable that the first and second estimated positions are standardized so as to have the same reliability.

The current position acquisition unit 15 acquires the current position of the moving body 1 by using the first and second estimated positions and the first and second reliability. For example, the current position acquisition unit 15 may set the estimated position corresponding to the higher reliability of the first and second reliability as the current position of the moving body 1. In this case, the current position acquisition unit 15 compares the first and second reliability, selects the estimated position corresponding to the higher reliability, and sets the selected estimated position as the current position of the moving body 1. May be. More specifically, when the first reliability is higher than the second reliability, the current position acquisition unit 15 sets the first estimated position corresponding to the first reliability as the current position. Will be done. In this case, the selection may be made in a situation where both the first and second estimated positions and the first and second reliability have been acquired in advance, and the first and second estimation positions may be made. It is also possible to calculate only the reliability of the above first and acquire the estimated position according to the larger reliability later. In this way, when one of the estimated positions is selected using the reliability and the selection target is switched according to the reliability, the current position acquired by the current position acquisition unit 15 suddenly becomes large. It is possible that it will change. Therefore, when the selection target is switched according to the reliability, and there is a difference of a predetermined threshold value or more between the current position before the selection target is switched and the current position after the selection target is switched. The current position may be acquired so as to smoothly connect from the current position immediately before switching to the newly selected current position. When the first and second reliability are the same, the current position acquisition unit 15 may use one of the estimated positions randomly selected as the current position of the moving body 1, or the first and first positions. The current position obtained by synthesizing the estimated positions of 2 may be acquired. The synthesis will be described later.

Further, the current position acquisition unit 15 may synthesize the current position of the moving body 1 from the first and second estimated positions by using, for example, the first and second reliability. The current position acquisition unit 15 may synthesize the first and second estimated positions by weighting the first and second reliability, for example. That is, the current position acquisition unit 15 may synthesize the current position by weighting addition of the first and second estimated positions using the first and second reliability as weights. Specifically, when the coordinate value of the current position is (X100, Y100), the current position acquisition unit 15 may perform the synthesis as follows.
X100 = (C1 x X101 + C2 x X102) / (C1 + C2)
Y100 = (C1 x Y101 + C2 x Y102) / (C1 + C2)
Here, the coordinate values of the first and second estimated positions are (X101, Y101) and (X102, Y102), respectively, and the first and second reliability are C1 and C2, respectively.

Further, the current position acquisition unit 15 may set the midpoint of the first and second estimated positions as the current position, for example, when the first and second reliability are both larger than a predetermined threshold value. .. The current position acquisition unit 15 may set the estimated position corresponding to the higher reliability as the current position, for example, when at least one of the first and second reliability is smaller than the threshold value. Further, the current position acquisition unit 15 may set the midpoint of the first and second estimated positions as the current position, for example, when the first and second reliability are equal.

The moving mechanism 16 moves the moving body 1. The moving mechanism 16 may or may not be capable of moving the moving body 1 in all directions, for example. Being able to move in all directions means being able to move in any direction. The moving mechanism 16 may have, for example, a traveling unit (for example, a wheel) and a driving means (for example, a motor, an engine, etc.) for driving the traveling unit. Further, the moving mechanism 16 may have a mechanism capable of acquiring the speed of a wheel or the like, for example, an encoder or the like. When the moving mechanism 16 can move the moving body 1 in all directions, the traveling portion may be an omnidirectional moving wheel (for example, an omni wheel, a mecanum wheel, or the like). As the moving mechanism 16, a known one can be used, and a detailed description thereof will be omitted.

The movement control unit 17 controls the movement of the moving body 1 by controlling the movement mechanism 16 using the current position acquired by the current position acquisition unit 15. The movement control may be control such as the movement direction of the moving body 1 and the start / stop of the movement. For example, when a movement route is set, the movement control unit 17 may control the movement mechanism 16 so that the moving body 1 moves along the movement route. More specifically, the movement control unit 17 may control the movement mechanism 16 so that the current position acquired by the current position acquisition unit 15 is along the movement path. Further, the movement control unit 17 may control the movement by using the map. Since the control of the movement mechanism 16 by the movement control unit 17 is known, detailed description thereof will be omitted.

Next, the operation of the moving body 1 will be described with reference to the flowchart of FIG.
(Step S101) The current position acquisition unit 15 determines whether or not to acquire the current position. Then, if the current position is acquired, the process proceeds to step S102, and if not, the process of step S101 is repeated until it is determined that the current position is acquired. The current position acquisition unit 15 may periodically determine that the current position is acquired, for example.

(Step S102) The first sensor 11 acquires information around the moving body 1. The information is, for example, distance measurement results in multiple directions.

(Step S103) The first position estimation unit 13 acquires the first estimated position and the first reliability by using the information acquired by the first sensor 11.

(Step S104) The second sensor 12 acquires information around the moving body 1. The acquisition of the information is, for example, the acquisition of a captured image.

(Step S105) The second position estimation unit 14 acquires the second estimated position and the second reliability by using the information acquired by the second sensor 12.

(Step S106) The current position acquisition unit 15 has the first estimated position and the first reliability acquired by the first position estimation unit 13, and the second estimation acquired by the second position estimation unit 14. The current position is obtained using the position and the second reliability. The current position may be, for example, either one of the first and second estimated positions, or may be a composite result of the first and second estimated positions. Then, the process returns to step S101.

Although not included in the flowchart of FIG. 2, it is assumed that the movement control unit 17 performs the movement control using the current position acquired by the current position acquisition unit 15. Further, the order of processing in the flowchart of FIG. 2 is an example, and the order of each step may be changed as long as the same result can be obtained. As described above, since it is preferable that the processes of steps S102 and S104 are performed synchronously, the processes of steps S102 and S103 and the processes of steps S104 and S105 may be performed in parallel. Further, in the flowchart of FIG. 2, the processing is terminated by the power off or the interrupt of the processing termination.

Next, the operation of the moving body 1 according to the present embodiment will be described with reference to a simple specific example. In this specific example, it is assumed that the moving body 1 is used as a transfer robot in the factory. Further, the current position acquisition unit 15 shall select the estimated position corresponding to the higher reliability as the current position. Further, the first reliability is acquired according to the degree of matching between the measurement result and the first map, and the second reliability is the specified feature points and the feature points included in the second map. It shall be acquired according to the degree of matching of.

First, it is assumed that the moving body 1 is moving in a busy area. In such a situation, the distance to the obstacle is measured by the first sensor 11, and the first estimated position and the first reliability are acquired by the first position estimation unit 13 (steps S101 to S103). .. The first reliability is a low value. This is because there are many obstacles (humans) that are not included in the first map. After that, the captured image is acquired by the second sensor 12, and the second estimated position and the second reliability are acquired by the second position estimation unit 14 (steps S104 and S105). Since the position estimation using the captured image is robust against new obstacles, the second reliability will not be as low as the first reliability. As a result, it is assumed that the second reliability is higher than the first reliability. Then, the current position acquisition unit 15 selects a second estimated position corresponding to the second reliability, which is a higher value, and passes the current position, which is the second estimated position, to the movement control unit 17 ( Step S106).

After that, the door of the factory was opened for people to come and go, and sunlight came into the factory through the door. Also, it is assumed that the second sensor 12 is facing the door at that time. Then, a part of the captured image acquired by the second sensor 12 becomes overexposed due to the influence of sunlight. As a result, it is assumed that the second reliability is greatly reduced and becomes smaller than the first reliability (steps S104 and S105). Then, the current position acquisition unit 15 selects the first estimated position according to the first reliability and switches to pass it to the movement control unit 17 as the current position (step S106). As described above, when there is a large difference between the current position before switching and the current position after switching, the current position acquisition unit 15 passes them to the movement control unit 17 so as to smoothly connect them. The current position may be gradually changed.

As described above, according to the moving body 1 according to the present embodiment, in order to acquire the current position of the moving body 1 using the first and second reliability, for example, the measurement result by the distance measuring sensor is used. If the first reliability corresponding to the first estimated position is lower than the second reliability corresponding to the second estimated position using the image captured by the image sensor, the second estimated position is currently set. By setting the position, it is possible to acquire the current position with higher accuracy. Further, it is possible to avoid using the result of the position estimation of the other automatically when the position estimation of one is inappropriate as in the above-mentioned conventional example. As a result, the accuracy of the current position becomes higher. Further, in the case where the first and second estimated positions are synthesized by the current position acquisition unit 15, and the first and second reliability are both high, the current position of the synthesis result is higher. It is considered to be highly accurate.

As described above, the accuracy of the measurement result of the distance measuring sensor decreases as the distance increases, so that the measurement result of a long distance may not be used. For example, as shown in FIG. 4, when the length to the measurement point acquired by the first sensor 11 which is the distance measuring sensor is larger than the predetermined threshold value L, the measurement point is self-determined. If the length to the measurement point is smaller than the threshold value L instead of being used for position estimation, the measurement point may be used for self-position estimation. If the distance to the measurement point is equal to the threshold value L, the measurement point may or may not be used for self-position estimation. In the situation shown in FIG. 4, the measurement point a will be used for self-position estimation by the first position estimation unit 13, and the measurement point b will be used for self-position estimation by the first position estimation unit 13. It will not be.

Also, even if the captured image is overexposed and overexposed, or underexposed and underexposed, it is unlikely that the entire captured image will be like that, and it will be partially white. Overexposure and underexposure often occur. Therefore, the second position estimation unit 14 acquires an appropriateness indicating the degree of appropriateness of the feature point for each feature point, and uses the appropriateness for each block of the captured image divided into a plurality of blocks. The block reliability is calculated, the second estimated position is acquired using the block with high block reliability, and the second reliability is calculated using the block reliability of the block used in the acquisition of the second estimated position. You may try to get it. For example, as shown in FIG. 5, when the captured image is divided into nine blocks 51 to 59 in advance, the second position estimation unit 14 may acquire the block reliability for each block. good. The block reliability may be, for example, a representative value of the appropriateness corresponding to the feature points included in the block. The representative value may be, for example, an average value, a median value, a maximum value, or the like. The appropriateness may be the similarity between the feature descriptor of the identified feature point and the feature descriptor of the feature point contained in the second map corresponding to the feature point (in this case, the degree of similarity). By tentatively matching the feature points specified in the captured image with the feature points included in the second map in block units, the correspondence between the feature points may be found), and the above evaluation values. It may be a value indicating the degree of appropriateness of other feature points. Further, the block having high block reliability may be, for example, a block whose block reliability exceeds a predetermined threshold value, and is a block having a predetermined number from the one with the highest block reliability. May be good. Further, to acquire the second estimated position by using the block having high block reliability means to acquire the second estimated position by using the feature points included in the block having high block reliability. Further, to acquire the second reliability by using the block reliability of the block used in the acquisition of the second estimated position means to obtain the block used in the acquisition of the second estimated position, that is, the block having high block reliability. The representative value of the corresponding block reliability may be the second reliability. The representative value may also be, for example, an average value, a median value, a maximum value, or the like, or may be a value calculated by weighting addition using the number of feature points included in the block as weights.

Further, in the above embodiment, each process or each function may be realized by centralized processing by a single device or a single system, or may be distributed processing by a plurality of devices or a plurality of systems. It may be realized by.

Further, in the above embodiment, the transfer of information performed between the components is performed by, for example, one of the components when the two components that transfer the information are physically different. It may be done by outputting information and accepting information by the other component, or if the two components that pass the information are physically the same, one component. It may be performed by moving from the processing phase corresponding to the other component to the processing phase corresponding to the other component.

Further, in the above embodiment, information related to the processing executed by each component, for example, information received, acquired, selected, generated, transmitted, or received by each component. Further, information such as threshold values, mathematical formulas, addresses, etc. used by each component in processing may be temporarily or for a long time held in a recording medium (not shown), even if it is not specified in the above description. In addition, each component or a storage unit (not shown) may store information on a recording medium (not shown). Further, the information may be read from the recording medium (not shown) by each component or a reading unit (not shown).

Further, in the above embodiment, when the information used in each component or the like, for example, the information such as the threshold value and the address used in the processing by each component and various setting values may be changed by the user, the above-mentioned The information may or may not be changed as appropriate by the user, even if it is not specified in the description. When the information can be changed by the user, the change is realized by, for example, a reception unit (not shown) that receives a change instruction from the user and a change unit (not shown) that changes the information in response to the change instruction. You may. The reception unit (not shown) may accept the change instruction from, for example, an input device, information transmitted via a communication line, or information read from a predetermined recording medium. ..

Further, in the above embodiment, when two or more components included in the mobile body 1 have a communication device, an input device, or the like, the two or more components may physically have a single device. , Or may have separate devices.

Further, in the above embodiment, each component may be configured by dedicated hardware, or a component that can be realized by software may be realized by executing a program. For example, each component can be realized by reading and executing a software program recorded on a recording medium such as a hard disk or a semiconductor memory by a program execution unit such as a CPU. At the time of execution, the program execution unit may execute the program while accessing the storage unit or the recording medium. Further, the program may be executed by being downloaded from a server or the like, or may be executed by reading a program recorded on a predetermined recording medium (for example, an optical disk, a magnetic disk, a semiconductor memory, etc.). good. Further, this program may be used as a program constituting a program product. Further, the number of computers that execute the program may be singular or plural. That is, centralized processing may be performed, or distributed processing may be performed.

Further, the present invention is not limited to the above embodiments, and various modifications can be made, and it goes without saying that these are also included in the scope of the present invention.

From the above, the moving body according to the present invention has the effect of being able to acquire the current position with high accuracy, and is useful as a moving body such as a transfer robot, for example.

1 Mobile body 11 1st sensor 12 2nd sensor 13 1st position estimation unit 14 2nd position estimation unit 15 Current position acquisition unit 16 Movement mechanism 17 Movement control unit

Claims (3)

It is a mobile body that moves autonomously.
A first sensor, which is a distance measuring sensor that measures the distance to an object around the moving object in a plurality of directions ,
A second sensor, which is an image sensor that acquires a captured image of the surroundings of the moving object,
The first estimated position and the first estimated position by estimating the position of the moving body using the distance measured by the first sensor and the first map showing the position of the obstacle in the moving area of the moving body. A first position estimation unit that acquires a first reliability corresponding to the first estimated position, and a first position estimation unit.
The position of the moving body is estimated by using the second map showing the position of the feature point specified in the captured image acquired by the second sensor and the position of the feature point in the moving region of the moving body. A second position estimation unit that acquires a second estimated position and a second reliability corresponding to the second estimated position, and a second position estimation unit.
A current position acquisition unit that acquires the current position of the moving body by using the first and second estimated positions and the first and second reliability.
A moving mechanism that moves the moving body and
A movement control unit that controls the movement mechanism using the current position is provided .
The second position estimation unit acquires an appropriateness indicating the degree of appropriateness of the feature point for each feature point, and uses the appropriateness to block each block of the captured image divided into a plurality of blocks. The reliability is calculated, the second estimated position is acquired using the block having the high block reliability, and the block reliability of the block used in the acquisition of the second estimated position is used to obtain the second estimated position. A moving body that gets a degree . The moving body according to claim 1 , wherein the current position acquisition unit sets an estimated position corresponding to the higher reliability of the first and second reliability as the current position of the moving body. The moving body according to claim 1 , wherein the current position acquisition unit synthesizes the current position of the moving body from the first and second estimated positions by using the first and second reliability.

Images