Determination device and determination method

Abstract

To provide a determination device and determination method, which enable accurate detection of a traffic mirror in an image.SOLUTION: A determination device provided herein comprises a detection unit and a determination unit. The detection unit detects areas with a given shape in multiple images captured at different times. When an optical flow that satisfies a given condition is present among the multiple images, the determination unit determines the areas to be a traffic mirror.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a determination device and a determination method. [Background technology]

[0002] Conventionally, there is known a technique for detecting a convex mirror from an image captured around a vehicle (for example, see Patent Document 1). In the conventional technique, a Hough transform or template matching is used to detect the circular or rectangular shape that constitutes the convex mirror.

[0003] Furthermore, because it is not possible to distinguish a convex mirror from road signs and other structures by simply detecting whether it is circular or rectangular, conventional techniques may use color information. For example, a convex mirror in an image is detected based on the fact that the outer edge of the convex mirror contains orange, and that the interior of the convex mirror contains few colors that are commonly used in road signs, such as red, blue, and white. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2016-110373 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, the conventional technology has a problem that it may not be possible to accurately detect the convex mirror in the image. For example, depending on the product of the convex mirror, the orange color of the outer edge may be small and may not appear in the image. In addition, the information of the outer edge is not always constant because it depends on the environment around the convex mirror.

[0006] The present invention has been made in consideration of the above, and has an object to provide a determination device and a determination method that can accurately detect convex mirrors in an image. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems and achieve the object, the determination device according to the present invention has a detection unit and a determination unit. The detection unit detects an area of ​​a predetermined shape that appears in a plurality of images captured at different times. The determination unit determines that the area is a convex mirror when an optical flow that satisfies a predetermined condition exists between the plurality of images. Effect of the Invention

[0008] According to the present invention, it is possible to accurately detect convex mirrors in an image. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of a determination device according to an embodiment. [Diagram 2] FIG. 2 is a diagram showing an example of optical flow. [Diagram 3] FIG. 3 is a diagram for explaining a method for determining a convex mirror. [Figure 4] FIG. 4 is a diagram showing an example of optical flow. [Diagram 5] FIG. 5 is a diagram showing an example of optical flow. [Figure 6] FIG. 6 is a flowchart showing the procedure of the process executed by the determination device according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, embodiments of a determination device and a determination method disclosed in the present application will be described in detail with reference to the accompanying drawings. Note that the present invention is not limited to the embodiments described below.

[0011] First, a determination device according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of the configuration of a determination device according to an embodiment.

[0012] The determination device 10 receives as input an image captured by a camera mounted on a vehicle.

[0013] Then, the determination device 10 determines whether or not a curved mirror is captured in the input image, and outputs the determination result.

[0014] The determination device 10 may be provided in a vehicle, or may be a server connected to an in-vehicle device provided in the vehicle via a network such as the Internet.

[0015] As shown in FIG. 1, the determination device 10 includes an input / output unit 11, a storage unit 12, and a control unit 13.

[0016] The input / output unit 11 is an interface for inputting and outputting data, and can receive input of an image from a camera or other device.

[0017] Furthermore, the input / output unit 11 outputs the determination result, for example, whether or not a curved mirror is shown in the input image, in the form of a text message, voice, binary data, or the like.

[0018] The control unit 13 and the memory unit 12 of the determination device 10 are realized by, for example, a computer having a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a flash memory, an input / output port, etc., or various circuits.

[0019] The storage unit 12 corresponds to a RAM or a flash memory. The RAM or the flash memory can store information about various programs executed by the determination device 10, and the like.

[0020] The determination device 10 may acquire the above-mentioned programs and various information via another computer or a portable recording medium connected via a wired or wireless network.

[0021] The CPU of the computer functions as the extraction unit 131, the detection unit 132, and the determination unit 133 of the control unit 13 by reading and executing a program stored in, for example, the ROM.

[0022] The extraction unit 131 extracts edges of an image. The extraction unit 131 can extract edges based on differences in pixel values ​​between pixels, using a known method such as object recognition.

[0023] The detection unit 132 detects an area having the shape of a convex mirror based on the edges extracted by the extraction unit 131.

[0024] For example, the detection unit 132 detects a region such as an ellipse or a rectangle that is set in advance as the shape of the convex mirror.

[0025] The detection unit 132 detects an area of ​​a predetermined shape that appears in a plurality of images captured at different times. In this manner, by using a plurality of images captured at different times, it becomes possible to detect an optical flow.

[0026] If there is an optical flow between a plurality of images that satisfies a predetermined condition, the determination unit 133 determines that the area is a convex mirror.

[0027] As a result, the determination device 10 of this embodiment can accurately detect a convex mirror in an image, regardless of the color of the outer edge of the convex mirror and the environment around the convex mirror.

[0028] In particular, when an optical flow including a downward component exists between a plurality of images, the determination unit 133 determines that the area is a curved mirror.

[0029] We explain why there is optical flow containing a downward component in the area of ​​the convex mirror.

[0030] Convex mirrors are sometimes installed at intersections, T-junctions, etc., where vehicles are about to enter, to allow drivers to see vehicles approaching from the left and right.

[0031] In this case, when viewed from a vehicle (hereinafter referred to as the vehicle itself) entering an intersection where a curved mirror is installed, an object approaching the intersection from the left or right will first be reflected in the upper left or upper right of the curved mirror, and then move to the lower right or lower left.

[0032] For example, when the vehicle is stationary, a moving object reflected in a curved mirror generates optical flow.

[0033] Furthermore, for the reasons mentioned above, optical flows that include downward components are thought to be generated by moving objects such as vehicles, so there is an increasing need to detect convex mirrors in order to avoid danger.

[0034] 2 is a diagram showing an example of an optical flow. As shown in FIG 2, an optical flow 11F is generated by a moving vehicle.

[0035] On the other hand, the optical flows represented by other dashed arrows are generated by trees and the like.

[0036] When the determination device 10 of this embodiment determines that a curved mirror is captured in the image, the determination result is used for processing such as risk prediction.

[0037] For this reason, it is important from the standpoint of danger prevention to detect convex mirrors that reflect moving objects such as vehicles.

[0038] In addition, by focusing on optical flow that includes a downward component, it is possible to prevent the movement of trees, etc. from being mistaken for the movement of a vehicle reflected in a curved mirror.

[0039] 3 is a diagram for explaining a method for determining whether a convex mirror is present. In the example of FIG. 3, each of a plurality of frames constituting a moving image is treated as an image by the determination device 10.

[0040] In the example of FIG. 3, it is assumed that the detection unit 132 detects the area surrounded by the dashed line as the area of ​​the convex mirror.

[0041] In the (n+1)th and (n+2)th frames, optical flow occurs within the region indicated by the dashed line, and therefore the determination unit 133 determines that this region is a convex mirror.

[0042] Furthermore, when the vehicle is moving, the determination device 10 can determine whether or not a curve mirror is reflected in the image using a method based on the length of the optical flow, in addition to or separately from the above determination method.

[0043] When the vehicle is traveling, structures such as buildings reflected in the curved mirror generate optical flow.

[0044] Fig. 4 is a diagram showing an example of optical flows. As shown in Fig. 4, among the optical flows generated by the structure in the area of ​​the convex mirror, the optical flow 22F in the center is shorter than the optical flows 21F and 23F in the peripheral areas, and the direction is different.

[0045] This is because a convex mirror is a convex mirror. On the other hand, a sign or the like is a flat surface with almost no unevenness, so unlike a convex mirror, there is no difference in the length and direction of the optical flow between the center and the periphery.

[0046] Therefore, when a plurality of optical flows with different lengths exist, the determination unit 133 determines that the area is a convex mirror.

[0047] In this case, the detection unit 132 detects an area of ​​a predetermined shape that appears in a plurality of images captured at different capture times by a moving vehicle.

[0048] This makes it possible to detect convex mirrors separately from signs and the like.

[0049] Furthermore, during the night or other times, the accuracy of edge extraction by the extraction unit 131 and the accuracy of area detection by the detection unit 132 are considered to be significantly reduced.

[0050] On the other hand, when the vehicle's headlights are turned on, the light from the headlights creates a light spot on the convex mirror.

[0051] Therefore, when no area is detected by the detection unit 132, but an optical flow of a specific direction caused by a light point at a predetermined height in the image exists, the judgment unit 133 can determine that the area of ​​the optical flow is a curved mirror.

[0052] Fig. 5 is a diagram showing an example of optical flow. As shown in Fig. 5, an optical flow 31F of a light spot generated by a headlight in the area of ​​a convex mirror is considered to include a downward component.

[0053] Therefore, by focusing on the light spots of the headlights, it is possible to determine whether or not a curved mirror is reflected in the image, even at night.

[0054] The procedure of the process executed in the determination device 10 according to the embodiment will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the procedure of the process executed by the determination device according to the embodiment.

[0055] 6, first, the extraction unit 131 extracts edges (step S101). Next, the detection unit 132 detects an area having the shape of a convex mirror based on the edges (step S102).

[0056] Here, the determination unit 133 determines whether or not there is an optical flow in a specific direction within the detected area (step S103).

[0057] If there is an optical flow in a specific direction (step S103: Yes), the determination unit 133 determines that the area is a curved mirror (step S104).

[0058] On the other hand, if there is no optical flow in a specific direction (step S103: No), the determination unit 133 determines that the area is not a curved mirror (step S105).

[0059] As described above, the determination device 10 according to the embodiment includes the detection unit 132 and the determination unit 133. The detection unit 132 detects an area of ​​a predetermined shape that appears in a plurality of images captured at different times. The determination unit 133 determines that the area is a convex mirror when an optical flow that satisfies a predetermined condition exists between the plurality of images.

[0060] In this way, the determination device 10 determines whether or not there is a convex mirror by focusing on the optical flow. Therefore, according to this embodiment, it is possible to accurately detect a convex mirror in an image, regardless of the color of the outer edge of the convex mirror and the environment around the convex mirror.

[0061] Here, the conventional technique will be described. In the conventional technique, a color space conversion is performed, for example, from YUV to RGB or HSV, for an area having the shape of a curved mirror detected from an image.

[0062] A histogram of color information is then generated in the converted color space, and if the frequency of hues near orange is high in the histogram, the area is determined to be a convex mirror.

[0063] As such, because conventional technology relies heavily on color analysis, it may not be able to make an accurate judgment when there is little orange on the outer edge of the convex mirror, when orange is not used in the convex mirror, or when the outer edge is not clearly visible in the image.

[0064] The determination method of this embodiment does not depend heavily on color, so it is possible to determine even curved mirrors that could not be determined using conventional technology.

[0065] In addition, the method of this embodiment and the conventional method may be implemented together. For example, when the conventional method cannot clearly determine whether or not a convex mirror is present, the method of this embodiment may be implemented.

[0066] Further advantages and modifications may readily occur to those skilled in the art. Thus, the invention in its broader aspects is not limited to the specific details and representative embodiments shown and described above. Accordingly, various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and equivalents thereof. [Explanation of symbols]

[0067] 10 Judgment device 11 Input / output section 12 Storage section 13 Control section 131 Extraction part 132 Detection unit 133 Judgment section

Claims

1. A detection unit that detects an area of ​​a predetermined shape that appears in a plurality of images captured at different times; A determination unit that determines that the area is a curved mirror when an optical flow that satisfies a predetermined condition exists between the plurality of images; having The detection unit detects an area of ​​a predetermined shape appearing in a plurality of images captured by a moving vehicle at different capture times; The determination unit determines that the area is a curved mirror when the optical flow of the central part of the area is shorter than the optical flow of the peripheral part and has a different direction. A determination device characterized by:

2. The determination device according to claim 1 , wherein the determination unit determines that the area is a curved mirror when an optical flow including a downward component exists between the plurality of images.

3. The determination device described in claim 1 or 2, characterized in that if the detection unit does not detect the area, and if there is an optical flow of a specific direction due to a light point at a predetermined height in the image, the determination unit determines that the area of ​​the optical flow is a curved mirror.

4. A determination method executed by a determination device, comprising: A detection step of detecting an area of ​​a predetermined shape appearing in a plurality of images captured at different times; A determination step of determining that the area is a curved mirror when an optical flow that satisfies a predetermined condition exists between the plurality of images; Including, The detection step detects an area of ​​a predetermined shape that appears in a plurality of images captured at different capture times by a moving vehicle; The determination step determines that the area is a curved mirror when the optical flow in the central part of the area is shorter and in a different direction than the optical flow in the peripheral part. A determination method comprising:

Images