Mobile body and guiding line detection apparatus

Abstract

To provide a guiding line detection apparatus which detects a guiding line while reducing the influence of external light.SOLUTION: A guiding line detection apparatus 2 includes: an image acquisition unit 11 which acquires a captured image obtained by imaging a front view of a mobile body 1 which moves along a guiding line provided on a floor surface and visually recognizable; an image conversion unit 12 which converts the captured image into an overhead view image; a grayscale conversion unit 13 which converts the converted captured image into a grayscale image; a filter application unit 14 which applies a filter for emphasizing the guiding line to the grayscale captured image; and a detection unit 15 which detects the guiding line in the filtered image. Accordingly, the guiding line is detected using the filtered image, thereby robustly detecting the guiding line while reducing the influence of external light.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a guiding line detection device that detects a visually recognizable guiding line provided on a floor surface, and a moving body that moves using the detected guiding line.

Background Art

[0002] Conventionally, in factories and the like, a guiding line provided on the floor surface has been detected, and a moving body has been moved along the detected guiding line. As a related technique, a luminance gradient vector of a photographed image of a road has been calculated, and a dividing line on the road has been detected using the luminance gradient vector (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, when using the luminance gradient vector, all lines with a difference in luminance are detected. For example, a shadow line formed by light entering through a window is also detected. Therefore, there is a problem that the target guiding line cannot be detected appropriately.

[0005] The present invention has been made to solve the above problems, and an object thereof is to provide a guiding line detection device that can detect a guiding line more robustly, and a moving body that moves using the detected guiding line.

Means for Solving the Problems

[0006] To achieve the above object, a guiding line detection device according to an aspect of the present invention includes an image acquisition unit that acquires a captured image of the front of a moving body that moves according to a visually recognizable guiding line provided on a floor surface, a grayscale conversion unit that converts the captured image into grayscale, a filter application unit that applies a filter for emphasizing the guiding line to the grayscale-converted captured image, and a detection unit that detects the guiding line in the image to which the filter has been applied.

Advantages of the Invention

[0007] According to a guiding line detection device or the like according to an aspect of the present invention, by detecting the guiding line using an image to which a filter for emphasizing the guiding line has been applied, the influence of external light or the like can be reduced, and the guiding line can be detected more robustly.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6A

Figure 6B

Figure 6C

Figure 6D

Figure 6E

Figure 6F

Figure 7A

Figure 7B

Figure 8

Figure 9

Figure 10

Embodiments for Carrying Out the Invention

[0009] Hereinafter, the mobile body and the induction line detection device according to the present invention will be described using embodiments. In the following embodiments, components and steps with the same reference numerals are the same or corresponding, and repeated descriptions may be omitted. The induction line detection device according to the present embodiment performs the detection of the induction line after applying a filter for emphasizing the induction line to the captured image after grayscale conversion.

[0010] FIG. 1 is a block diagram showing the configuration of the mobile body 1 according to the present embodiment. The mobile body 1 according to the present embodiment moves according to a visually recognizable induction line provided on the floor surface, and includes an induction line detection device 2, a moving mechanism 16 for moving the mobile body 1, and a movement control unit 17 for controlling the moving mechanism 16 using the detection result of the induction line by the induction line detection device 2. The induction line detection device 2 includes an image acquisition unit 11, an image conversion unit 12, a grayscale conversion unit 13, a filter application unit 14, and a detection unit 15. The use of the mobile body 1 is not particularly limited, but the mobile body 1 may be, for example, a mobile body for carrying, or a mobile body for other uses such as security and cleaning.

[0011] It is assumed that a guiding line is provided on the floor surface of the moving environment of the moving body 1. The floor surface of the moving environment is preferably horizontal normally. The guiding line may be provided, for example, to guide the moving body 1, or may be a demarcation line or the like for partitioning other uses, such as a passage for people or a traveling area for a forklift. The guiding line may be, for example, a tape-like object such as vinyl tape pasted on the floor surface, or may be provided by applying paint to the floor surface. The guiding line is preferably a line with a certain width, for example. The guiding line is visually recognizable. That is, the guiding line can be detected in the image. Since the guiding line is visually recognizable, it is preferably a color different from that of the floor surface. The guiding line may be, for example, a line of a specific color determined in advance. Also, the guiding line may be, for example, a straight line or a curve. In the present embodiment, the case where the guiding line is a straight line will be mainly described.

[0012] FIG. 2 is a plan view showing an example of the situation in a factory where a guiding line 5 is provided on the floor surface. In FIG. 2, it is assumed that the guiding line 5 is a demarcation line for partitioning a passage for people near the arrangement 7 in the factory and a traveling area for a forklift.

[0013] The image acquisition unit 11 acquires a captured image of the front of the moving body 1. The image acquisition unit 11 may be, for example, an optical device such as a camera that captures an image, or may be a device that acquires an image captured by an optical device such as a camera. In the present embodiment, the case where the image acquisition unit 11 is a camera will be mainly described. The captured image is preferably a color image. The image acquisition unit 11 preferably repeats the acquisition of the captured image. The image acquisition unit 11 may repeat the acquisition of the captured image, for example, periodically or irregularly. Also, the image acquisition unit 11 may acquire a moving image, for example. In this case, one frame constituting the moving image may be considered as the captured image.

[0014] The optical axis of the camera for taking a captured image preferably generally faces forward in the traveling direction of the moving body 1. The optical axis may, for example, extend in the horizontal direction, or may face the floor side such that the depression angle is a positive value. Even in the latter case, it is preferable that a region far in the traveling direction is also included in the captured image. Also, for example, the captured image may be taken such that the left-right direction of the captured image is the horizontal direction of the real space. FIG. 3 is a diagram showing an example of a captured image taken when the moving body 1 exists at the position shown in FIG. 2. As shown in FIG. 3, the captured image becomes an image with a perspective. Also, due to the external light entering through the window 3, a shadow is formed on the floor surface. In FIG. 3, the shadow is indicated by hatching.

[0015] The image conversion unit 12 converts the captured image acquired by the image acquisition unit 11 into a captured image viewed from above. Note that the converted captured image may also be referred to as a planar image. The planar image, which is a captured image viewed from above, may be, for example, an image viewed from a direction perpendicular to the floor surface, for example, the vertical direction. The image conversion unit 12 may, for example, convert the captured image into a planar image by a homography transformation. Note that a method for converting a captured image into a planar image is known, and a detailed description thereof is omitted. FIG. 4 is a diagram showing an example of a planar image obtained by converting the captured image shown in FIG. 3. As shown in FIG. 4, the planar image becomes an image without a perspective. The image conversion unit 12 may, for example, repeatedly convert each of the acquired captured images into a planar image, which is a captured image viewed from above.

[0016] The grayscale conversion unit 13 converts the captured image into grayscale. As an example, the grayscale conversion unit 13 may perform grayscale conversion on the captured image after conversion by the image conversion unit 12. This grayscale conversion may be performed, for example, such that the luminance of the color image becomes the luminance of the grayscale image. Further, the grayscale conversion unit 13 may perform grayscale conversion such that, for example, the difference from the reference color, which is the color of the guiding line, becomes larger as the difference in color from the reference color becomes larger. That is, when converting a pixel of a certain color to grayscale, the closer the color is to the reference color, the closer the value of the pixel after grayscale conversion is to the value of the pixel of the reference color after grayscale conversion, and the farther the color is from the reference color, the farther the value of the pixel after grayscale conversion is from the value of the pixel of the reference color after grayscale conversion. For example, when converting to 8-bit grayscale, the value of the pixel of the reference color may be set to 255, and the grayscale conversion may be performed such that the value becomes closer to 0 as the difference in color from the reference color becomes larger. By doing so, it becomes possible to more appropriately detect the guiding line in the grayscale planar image after conversion. As in the above example, grayscale conversion may be performed such that pixels closer to the reference color have larger luminance values. In this case, for example, the guiding line of the reference color is converted to white by grayscale conversion. In the present embodiment, this case will be mainly described. FIG. 5 is a diagram showing an example of a captured image obtained by converting the planar image shown in FIG. 4 into grayscale. For example, as shown in FIG. 5, in the captured image after grayscale conversion, the luminance value of the guiding line 5 may be the largest value. The grayscale conversion unit 13 may, for example, convert each of the repeatedly acquired planar images into a grayscale image.

[0017] Note that, regardless of the method of performing grayscale conversion, the greater the color difference from the reference color, the greater the difference from the pixel value of the reference color. For example, in a color space such as RGB or CMY, the distance from the reference color of the guiding line may be specified, and the value after conversion may be determined according to the specified distance. That is, the greater the specified distance, the greater the difference from the pixel value of the reference color, and grayscale conversion may be performed. Also, for example, in an HSV color space or an xyz color space, the distances from each component of the reference color of the guiding line (for example, hue, saturation, brightness, etc. in the HSV color space) may be specified respectively, and the value after conversion may be determined according to the distances of the specified components. In this case, for example, one distance is obtained using the distances of the components, and the greater the one distance, the greater the difference from the pixel value of the reference color, and grayscale conversion may be performed. The one distance may be obtained, for example, by multiplying or adding the distances of the components.

[0018] Note that the order of performing image conversion by the image conversion unit 12 and grayscale conversion by the grayscale conversion unit 13 on the captured image is not limited. For example, grayscale conversion may be performed after converting the captured image into a planar image, or the captured image may be converted into a planar image after grayscale conversion. In the present embodiment, the former case will be mainly described.

[0019] The filter application unit 14 applies a filter for emphasizing the guiding line to the captured image on which image conversion by the image conversion unit 12 and grayscale conversion by the grayscale conversion unit 13 have been performed. Details of the processing related to the application of this filter will be described later.

[0020] The detection unit 15 detects the guiding line in the image to which the filter is applied. As an example, the detection unit 15 may detect the guiding line by a Hough transform in the image to which the filter is applied. For example, when the guiding line is a straight line, the detection unit 15 may perform a Hough transform to detect the straight line. Also, the detection unit 15 may detect the guiding line by template matching or feature point extraction. As an example, when the guiding line is a straight line, the detection unit 15 may detect the guiding line by using a straight-line template image. Also, when the feature point extraction is performed, the detection unit 15 may extract feature points such as SIFT key points or SURF key points, and detect the guiding line including the feature points. The detection result by the detection unit 15 may be, for example, information indicating the position of the guiding line in the image.

[0021] In addition, when the width of the guiding line is large in the image to which the filter is applied, the detection unit 15 may detect the guiding line after performing a thinning process on the image to which the filter is applied, for example. This is because thinner lines are more suitable for detection. The thinning process may be, for example, an image process in which a line having a width is thinned by narrowing the line width, or may be edge detection. As the former thinning process, for example, Tamura, Hilditch, Zhang-Suen, and other algorithms that thin an image after binarizing it are known. In the latter edge detection, edges on both sides of the width direction of the guiding line are usually detected, but for example, edge detection may be performed so that only one edge is left. In order to leave only one edge, for example, when detecting an edge using a differential value such as a luminance value, only pixels whose differential value is greater than a positive threshold value may be detected as edges, and pixels whose differential value is less than a negative threshold value may not be detected as edges.

[0022] The moving mechanism 16 moves the moving body 1. In the present embodiment, the case where the moving mechanism 16 is a mechanism for running the moving body 1 on the floor surface will be mainly described. The moving mechanism 16 may be, for example, one that can move the moving body 1 in all directions, or may not be so. Being able to move in all directions means being able to move in an arbitrary direction. The moving mechanism 16 may have, for example, a traveling unit (such as wheels) and a driving means (such as a motor or an engine) for driving the traveling unit. In the case where the moving mechanism 16 can move the moving body 1 in all directions, the traveling unit may be an omnidirectional wheel (such as an omni wheel or a mecanum wheel). Since a known moving mechanism 16 can be used, a detailed description thereof will be omitted.

[0023] The movement control unit 17 controls the moving mechanism 16 so that the moving body 1 moves according to the detected guiding line. For the moving body 1 to move according to the guiding line, for example, the moving body 1 may move along the guiding line. When the position of the guiding line in the image is specified by the detection unit 15, the position of the guiding line in the local coordinate system of the moving body 1 can be specified using the specification result. Therefore, the movement control unit 17 may control the moving mechanism 16 using the position of the guiding line in the local coordinate system of the moving body 1 so that movement according to the guiding line is performed.

[0024] How the mobile body 1 moves along the guiding line may be set in advance. Then, the movement control unit 17 may control the movement according to the guiding line by using, for example, the movement distance obtained by using an encoder or the like provided in the movement mechanism 16 according to the setting. For example, the movement control unit 17 may move the mobile body 1 along the guiding line by a predetermined distance, change the traveling direction of the mobile body 1 to a predetermined direction at the branching point of the guiding line after the movement, and then repeat the process of moving the mobile body 1 along the guiding line by a predetermined distance again. Further, the movement control unit 17 may stop the mobile body 1 at a predetermined position. And at that position, for example, loading and unloading of the object to be conveyed may be performed.

[0025] Also, when a marker is arranged in the movement area of the mobile body 1, the movement control unit 17 may perform movement control using the marker. The marker may indicate, for example, the stop position of the mobile body 1 or the traveling direction of the mobile body 1 at the branching point of the guiding line. In this case, the movement control unit 17 may detect the marker by using template matching or the like in the captured image before the application of the filter, for example, the captured image which is a planar image, and perform movement control using the detected marker. As an example, the marker may be arranged on the guiding line.

[0026] Movement along the guiding line may be, for example, moving along the guiding line on the guiding line. Further, movement along the guiding line may be moving along a virtual line parallel to the guiding line on that virtual line. The virtual line does not exist in the moving area of the moving body 1, that is, it is a line that cannot be visually observed in the real environment. In this case, the moving body 1 can move parallel to the guiding line at a position different from the guiding line. For example, even when using, as the guiding line, a partition line provided near an arrangement in a factory or the like, by causing the moving body 1 to move on a virtual line that is moved parallel in a direction away from the arrangement along the guiding line, the possibility of the moving body 1 coming into contact with the arrangement can be reduced. As an example, the distance between the guiding line and the virtual line may be set in advance.

[0027] Next, the application of the filter to the captured image will be described. FIG. 6A is a diagram for explaining the application of the filter. In FIG. 6A, for simplicity of explanation, a 1-row and 5-column filter is used. Also, in FIG. 6A, only the values of some pixels of the image are shown. The image before the application of the filter is a grayscale planar image. As shown in FIG. 6A, the filter application unit 14 identifies, in the image before the application of the filter, a region of the same size as the filter, that is, the region of elements B1 to B5, and calculates the sum of the products of the respective elements of the filter (for example, A1, A2, etc.) and the respective elements of the identified region (for example, B1, B2, etc.). As shown in FIG. 6A, assuming that the calculation result is X1, the filter application unit 14 sets the value of the element corresponding to the identified region in the image after the application of the filter to X1. When the calculation result of the sum of the products of the respective elements of the filter and the identified region becomes a negative value, for example, the calculation result may be set to 0. This is because the value of each pixel usually becomes a value of 0 or more. Next, the filter application unit 14 then shifts the region of the object to be identified by one pixel to the right in the image before the application of the filter, identifies the region of elements B2 to B6, and calculates the sum of the products of the respective elements of the filter and the respective elements of the identified region. Assuming that the calculation result is X2, the filter application unit 14 sets the value of the element at the position shifted in the same manner as the region of the object to be identified with respect to the element calculated previously in the image after the application of the filter, that is, the element on the right side of X1, to X2. In this way, by repeating the same process while shifting the region of the image to which the filter is applied, the image after the application of the filter is generated.

[0028] Note that the process of applying this filter is the same as the process of applying a filter in the convolutional layer of a convolutional neural network, that is, the process of convolution using the filter. It is preferable that the application of the filter is performed while shifting the area to which the filter is applied by one pixel at a time. That is, in the application of the filter, the stride may be "1". Also, when such a filter is applied, the size of the image after the application of the filter becomes smaller than the size of the image before the application of the filter. When it is desired to make the sizes of the images the same before and after the application of the filter, padding may be performed as appropriate. The padding is not particularly limited, and for example, zero padding, mirror padding, repetitive padding, etc. may be used.

[0029] Next, a filter for emphasizing the induction line will be described. This filter may have one or more rows. Here, first, a one-row filter will be described, and then a filter having two or more rows will be described. FIG. 6B is a diagram showing a one-row filter. As shown in FIG. 6B, the filter may have, in each row, a first region corresponding to the induction line, two second regions provided on both sides of the first region, and two third regions provided between the first region and the two second regions. Each value in the first region may be positive. Also, each value in the second region may be negative. Also, each value in the third region may be 0.

[0030] Since the image to which the filter is applied is an image used for movement according to the guidance line, usually, the guidance line extends in the vertical direction in that image. Also, the direction along the rows of the filter, that is, the left - right direction in FIG. 6B, is the left - right direction in the image to which the filter is applied. Therefore, the direction along the rows of the filter is orthogonal to the direction in which the guidance line extends. Thus, the first region corresponding to the guidance line may be set, for example, to correspond to the line width of the guidance line. Also, the second region is preferably set within a range that does not cover the guidance line adjacent to the guidance line used by the moving body 1 for movement. That is, it is preferable that the second region is set so that the widths of the second and third regions, that is, the actual distance corresponding to the length along the rows, are shorter than the shortest distance between adjacent guidance lines in the movement region of the moving body 1. Also, it is preferable that the widths of each of the two second regions are equal.

[0031] The third region is set to be able to cope even when the width of the guidance line fluctuates. For example, as described above, in the image to which the filter is applied, it is ideal for the guidance line to extend in the vertical direction, but in reality, it is also conceivable that the guidance line has a slight angle with respect to the vertical direction. Even in such a case, the third region may be provided so as to enable appropriate detection of the guidance line. It is preferable that the widths of each of the two third regions are equal. Note that in the filter, the third region may not be provided. In this case, in each row of the filter, two second regions may exist on both sides of the first region without any interval. In this way, the two second regions provided on both sides of the first region may be provided adjacent to the first region, or may be provided via the third region.

[0032] Note that, as an example, each value in the first region of the filter may be set such that the sum of each value included in the first region is 1. Also, each value in the first region may, for example, be the same, or not. Further, as an example, each value in the two second regions of the filter may be set such that the sum of each value included in the first region and the second regions is 0. Each value in the second region may, for example, be the same, or not.

[0033] Here, an example of applying the filter to an image will be described. Assume that FIG. 7A is an image before applying the filter. The image shown in this FIG. 7A is an 8-bit grayscale planar image and includes a guiding line extending in the vertical direction. In FIG. 7A, only the guiding line is white (value is 255), and the area other than the guiding line is black (value is 0). If the filter to be applied is the one shown in FIG. 6C, the image after applying the filter will be as shown in FIG. 7B. Note that zero-padding is performed in applying the filter. In this way, by applying the filter, the white color of the guiding line is maintained, but the white noise with a width narrower than the guiding line will not have its white color maintained. Therefore, the noise can be removed by applying the filter, and only the guiding line can be robustly detected. Also, by applying the filter, the guiding line having a width of 5 pixels is substantially converted to a width of 3 pixels, so applying the filter also has the effect of thinning the guiding line.

[0034] Note that, in the filters shown in FIGS. 6B and 6C, each value in the first region is the same, but it may not be so. For example, as shown in FIG. 6D, each value in the first region may be set such that the value corresponding to the center in the width direction of the guiding line is the largest, and the value becomes smaller as going to both end sides in the width direction of the guiding line. Also in this case, it is preferable that each value is set such that the sum of all values in the first region is 1.

[0035] Also, as shown in FIG. 6E, the filter may have two or more rows. By applying such a filter having two or more rows to an image, smoothing can also be performed in the vertical direction of the captured image. For example, even when the guiding line is wavy, in the image after applying the filter, the guiding line can be made closer to a straight line. Therefore, by using a filter having two or more rows, a straight guiding line extending in the vertical direction can be emphasized in the image to which the filter is applied. In a filter having two or more rows, in each row, the widths of the first to third regions are usually the same. Each row included in the filter may all be the same, or may not be. In the former case, as shown in FIG. 6E, the values of each column, that is, the values arranged in the vertical direction, may all be the same. In the latter case, for example, as shown in FIG. 6F, the values of each column may be such that the values of the first row and the third row are the same, and the value of the second row is twice the values of the first row and the third row. Thus, when the filter has a plurality of rows, in each column, the values may be set such that the absolute value of the value of the central row is the largest, and the absolute value of the value becomes smaller as going to the topmost row and the bottommost row. In this case, smoothing that emphasizes the rows near the center can be performed. Note that when the filter has two or more rows, the values of the first region in each row may, for example, be the same (see FIGS. 6E and 6F), or may not be. When the filter has two or more rows, the values of the second region in each row may, for example, be the same.

[0036] FIG. 8 is a diagram showing an example of an image after applying the filter of FIG. 6C to the grayscale planar image shown in FIG. 5. For example, as shown in FIG. 8, in the captured image after applying the filter, the guiding line 5 is shown as a white line having a predetermined width extending in the vertical direction. The filter application unit 14 may, for example, repeatedly perform the process of applying the filter to each of the repeatedly acquired grayscale planar images.

[0037] FIG. 9 is a diagram showing the experimental results of performing each process on an actual photographed image. In FIG. 9, the left photograph shows the photographed image viewed from above after image conversion by the image conversion unit 12, the middle photograph shows the photographed image after grayscale conversion by the grayscale conversion unit 13, and the right photograph shows the image to which the filter has been applied. As shown in the right photograph of FIG. 9, it can be seen that the guiding lines that are difficult to identify in the photographed image, which is a planar image, can be easily identified in the image to which the filter has been applied.

[0038] Next, the operation of the guiding line detection device 2 will be described using the flowchart of FIG. 10. (Step S101) The image acquisition unit 11 determines whether to acquire a photographed image. If it is determined to acquire a photographed image, the process proceeds to step S102. Otherwise, the process of step S101 is repeated until it is determined to acquire a photographed image. The image acquisition unit 11 may, for example, periodically determine whether to acquire a photographed image.

[0039] (Step S102) The image acquisition unit 11 acquires a photographed image. The image acquisition unit 11 may, for example, acquire a photographed image by taking a photograph, or may acquire a photographed image by receiving a photographed image. The photographed image may be stored in a recording medium (not shown) or the like.

[0040] (Step S103) The image conversion unit 12 converts the photographed image into a planar image, which is the photographed image viewed from above. Note that the image conversion unit 12 may convert all the acquired photographed images into planar images, or may convert some of the photographed images into planar images. This is because when photographed images are acquired frequently, it is not necessarily required to convert all the photographed images into planar images. The planar image may be stored in a recording medium (not shown) or the like.

[0041] (Step S104) The grayscale conversion unit 13 performs grayscale conversion on the planar image.

[0042] (Step S105) The filter application unit 14 applies a filter to the captured image that has undergone image conversion by the image conversion unit 12 and grayscale conversion by the grayscale conversion unit 13.

[0043] (Step S106) The detection unit 15 detects the guiding line in the image to which the filter has been applied. Information indicating the position of the detected guiding line may be passed to, for example, the movement control unit 17. Then, the process returns to Step S101.

[0044] Although not included in the flowchart of FIG. 10, the movement control unit 17 may control the movement mechanism 16 so that the moving body 1 moves along the detected guiding line. Also, the order of the processes in the flowchart of FIG. 10 is an example, and if the same result can be obtained, the order of each step may be changed. For example, after grayscale conversion, image conversion to a planar image may be performed. Also, in the flowchart of FIG. 10, the process ends due to a power-off or a processing end interrupt. For example, when the moving body 1 arrives at the destination, the process of the flowchart of FIG. 10 may end.

[0045] Next, the operation of the moving body 1 according to the present embodiment will be described using a specific example. In this specific example, it is assumed that the moving body 1 moves according to the guiding line 5 in the factory shown in FIG. 2. Also, it is assumed that the movement control unit 17 performs movement control according to the route along each guiding line from the starting point to the destination, which is stored in a recording medium (not shown) in advance.

[0046] When the moving body 1 starts moving, first, the image acquisition unit 11 acquires a captured image and passes the acquired captured image to the image conversion unit 12 (Steps S101, S102). Assume that the captured image is, for example, the one shown in FIG. 3. Upon receiving the captured image, the image conversion unit 12 converts the captured image into a planar image as seen from above and passes it to the grayscale conversion unit 13 (Step S103). The planar image is, for example, the one shown in FIG. 4.

[0047] When receiving a planar image, the grayscale conversion unit 13 performs grayscale conversion such that the difference from the pixel value of the reference color becomes larger as the color difference from the reference color, which is the color of the guiding line, becomes larger, and passes the captured image after the grayscale conversion to the filter application unit 14 (step S104). The captured image after the grayscale conversion is, for example, as shown in FIG. 5.

[0048] When receiving the captured image after the grayscale conversion, the filter application unit 14 performs zero-padding at the boundary portions at the left end and the right end of the captured image, applies the filter shown in FIG. 6C to the captured image after the padding, and passes the image after the application of the filter to the detection unit 15 (step S105). The image after the application of the filter is, for example, as shown in FIG. 8.

[0049] When receiving the image after the application of the filter, the detection unit 15 detects the guiding line and passes the detection result of the guiding line to the movement control unit 17 (step S106). In FIG. 8, there are two guiding lines 5 extending in the traveling direction (vertical direction), but the detection result of the left guiding line 5 existing at a position closer to the center in the left-right direction of the planar image may be passed to the movement control unit 17. When receiving the detection result of the guiding line, the movement control unit 17 controls the movement mechanism 16 so that the moving body 1 moves according to the detected guiding line. As described above, the movement control unit 17 may appropriately change the moving direction or stop the moving body 1. In this way, by repeating processes such as acquisition of the captured image, conversion of the captured image, application of the filter to the captured image after the conversion, detection of the guiding line in the image after the application of the filter, and movement control according to the detected guiding line, the moving body 1 can move to the destination along the guiding line.

[0050] As described above, according to the mobile body 1 according to the present embodiment, by detecting the guiding line using the image to which the filter for emphasizing the guiding line is applied, it becomes possible to detect the guiding line more robustly. For example, even in a situation where a shadow is formed on the floor surface by external light entering through a window or in a situation where an object of the same color system as the guiding line exists in the moving space of the mobile body 1, it is possible to realize more stable detection of the guiding line. As a result, it becomes possible to more surely realize the movement of the mobile body 1 according to the guiding line. Further, by applying a filter to the captured image after the image conversion by the image conversion unit 12, the same filter can be applied to the entire captured image.

[0051] In the present embodiment, the case where a filter is applied to a captured image, which is a planar image subjected to image conversion by the image conversion unit 12, has been mainly described, but this is not essential. The filter may be applied to the captured image before it is converted into a planar image. In this case, for example, grayscale conversion by the grayscale conversion unit 13 may be performed, and the filter may be applied to a captured image in which image conversion by the image conversion unit 12 has not been performed. In a captured image that is not a planar image, as shown in FIG. 3, the line width of the guiding line changes according to the vertical position of the captured image. Therefore, when the filter application unit 14 applies a filter to a captured image that is not a planar image, it may apply a filter corresponding to the line width corresponding to that position for each vertical position of the captured image. That is, the filter applied to a captured image that is not a planar image may have the widths of the first to third regions corresponding to the height of the captured image (i.e., the position in the vertical direction of the image) to which the filter is applied. When the width of the guiding line provided on the floor surface is determined and the relative positional relationship between the guiding line and the moving body 1 during the movement of the moving body 1 is determined, in a captured image that is not a planar image, the relationship between the vertical position of the image and the line width of the guiding line included in the image is determined. Therefore, for example, filters may be prepared for each vertical position of the image such that the width of the first region is equal to the line width of the guiding line and applied to the image. Alternatively, such a filter may be dynamically generated according to the vertical position of the image and applied to the image. Further, when applying a filter having two or more rows to the image, in the filter, for example, the widths of the first to third regions may be different for each row. Note that the width of the region may be the length of the region in each row. Also, in this case as well, the filter may not have the third region. A filter that does not have the third region may have the widths of the first and second regions corresponding to the height of the captured image to which the filter is applied.

[0052] Also, in this embodiment, the case where the optical axis of the camera that captures the captured image is parallel to the guiding line in a plan view has been mainly described, but it does not have to be so. In a plan view, the movement may be performed with the optical axis of the camera having a predetermined angle with the guiding line. For example, when the moving body 1 can move in all directions, such movement is also possible. In this case, for example, in a plan view, when the angle formed by the direction of the guiding line and the direction of the optical axis of the camera is θ, that is, when the angle formed by the direction of the guiding line and the vertical direction of the captured image (which is a planar image) is θ, the widths of the first to third regions in each row of the filter may be set to the widths described above multiplied by 1 / cosθ. In addition, when the movement is being performed so that the direction of the guiding line is in the vertical direction in the captured image (which is a planar image), even when the direction of the guiding line deviates from the vertical direction due to some factor, the widths of the respective regions in each row of the filter may be changed in the same manner.

[0053] Also, in this embodiment, the case where the guiding line is a straight line has been mainly described, but the guiding line may be a curve. The curved guiding line may be, for example, a curved guiding line with a constant curvature. Also in this case, the detection unit 15 can detect, for example, a curved guiding line with a constant curvature by means of a Hough transform.

[0054] Also, when detecting the guiding line using a Hough transform, the result of the Hough transform may sometimes fluctuate. In order to avoid such a situation, processing for stabilizing the result may be performed. The processing for stabilizing the result of the Hough transform may be, for example, filtering using a Kalman filter or the like, or may be performing weighted addition or the like for excluding outliers or reducing the influence of outliers.

[0055] Also, when a plurality of guiding lines are included in a planar image, the moving body 1 may also detect guiding lines other than the guiding line currently used for movement. As a result, when the moving body 1 is moving according to a certain guiding line, there is a possibility that it may erroneously move according to a different guiding line. To avoid such inappropriate movement, the detection unit 14 may perform detection of guiding lines only in the vicinity of the guiding line used in the current movement.

[0056] Also, when the moving body 1 is movable in a plurality of directions like an omnidirectional moving cart, for example, the image acquisition unit 11 may be able to acquire a captured image in front of the moving direction for each of the plurality of moving directions. In this case, for example, in the moving body 1, cameras for capturing captured images for each of the plurality of directions may be provided. Then, the image acquisition unit 11 may acquire a captured image in front of the moving direction corresponding to the current moving direction.

[0057] Also, in the above embodiment, the case where the guiding line detection device 2 is a stand-alone device has been described. However, the guiding line detection device 2 may be a stand-alone device or may be a server device in a server-client system. In the latter case, the image acquisition unit 11 may receive a captured image captured in the moving body 1 from the moving body 1. Also, information regarding the guiding line detected by the detection unit 15 may be transmitted to the moving body 1 by a transmission unit (not shown) included in the guiding line detection device 2. In this case, the moving body 1 may move according to the guiding line detected by the guiding line detection device 2 which is a server device.

[0058] Also, in the above embodiment, each process or each function may be realized by being centrally processed by a single device or a single system, or may be realized by being distributively processed by a plurality of devices or a plurality of systems.

[0059] In the above-described embodiments, each component may be configured by dedicated hardware, or components that can be realized by software may be realized by executing a program. For example, each component can be realized by a program execution unit such as a CPU reading and executing a software program recorded on a recording medium such as a hard disk or a semiconductor memory. At the time of its execution, the program execution unit may execute the program while accessing a storage unit or a 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.). Also, this program may be used as a program constituting a program product. Further, the computer that executes the program may be singular or plural. That is, centralized processing may be performed, or distributed processing may be performed.

[0060] Moreover, the above embodiments are examples for specifically implementing the present invention and do not limit the technical scope of the present invention. The technical scope of the present invention is indicated by the claims rather than the description of the embodiments, and it is intended that changes within the literal scope of the claims and the scope of equivalent meaning are included.

Description of Reference Numerals

[0061] 1 Moving body, 2 Inductive line detection device, 5 Inductive line, 11 Image acquisition unit, 12 Image conversion unit, 13 Grayscale conversion unit, 14 Filter application unit, 15 Detection unit, 16 Moving mechanism, 17 Movement control unit

Claims

1. An image acquisition unit that acquires a captured image by photographing the front of a moving body that moves according to a visually recognizable guiding line provided on a floor surface; A grayscale conversion unit that converts the captured image into grayscale; A filter application unit that applies a filter for emphasizing the guiding line to the grayscale-converted captured image; A detection unit that detects a guiding line in the image to which the filter has been applied, and a guiding line detection device comprising the same.

2. Further comprising an image conversion unit that converts the captured image into a captured image viewed from above, The filter application unit applies a filter to the captured image that has been grayscale-converted by the grayscale conversion unit and image-converted by the image conversion unit. The guiding line detection device according to claim 1.

3. The filter has one or more rows, and in each row, has a first region corresponding to a guiding line and two second regions respectively provided on both sides of the first region, Each value in the first region is positive, Each value in the second region is negative. The guiding line detection device according to claim 2.

4. The filter has two or more rows. The guiding line detection device according to claim 3.

5. The filter has, in each row, two third regions respectively provided between the first region and the two second regions, Each value in the third region is 0. The guiding line detection device according to claim 3.

6. The filter has one or more rows, and in each row, has a first region corresponding to a guiding line and two second regions respectively provided on both sides of the first region, Each value in the first region is positive, Each value in the second region is negative, The filter has widths of the first and second regions corresponding to the height of the captured image to which the filter is applied. The guiding line detection device according to claim 1.

7. The grayscale conversion unit performs grayscale conversion such that the difference from the value of the pixel of the reference color becomes greater as the color difference from the reference color, which is the color of the guiding line, becomes greater. The guiding line detection device according to any one of claims 1 to 6.

8. A guiding line detection device according to any one of claims 1 to 6, A moving mechanism that moves the moving body, A moving control unit that controls the moving mechanism so that the moving body moves according to the guiding line detected by the detection unit, and a moving body comprising the same.

Images