JP2019175107A - Recognition device, recognition method, program, and data generation device - Google Patents

Abstract

To provide a technology capable of executing robust discrimination against climate while reducing computation amounts required to discrimination based on a photographed image.SOLUTION: Provided is a recognition device, and that comprises: an image acquisition section for acquiring a photographed image captured by a camera; a data generation section for generating reduced high frequency information by extracting and reducing high frequency components based on first luminance information of the photographed image; and a discriminator for performing prescribed discrimination based on the reduced high frequency information.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a recognition device, a recognition method, a program, and a data generation device.

  In recent years, techniques for performing various types of identification based on captured images captured by a camera are known. For example, a technology is disclosed in which the installation position of a license plate is specified based on a captured image and information (for example, a mark, a character string, etc.) described on the license plate is recognized based on the specified installation position. (For example, refer to Patent Document 1).

JP 2007-299144 A

  When the identification is performed by the classifier, the luminance information itself of the captured image is input to the classifier. However, when the luminance information itself of the captured image is input to the discriminator, the amount of calculation required for discrimination tends to increase. As an example, when a classifier based on deep learning (deep learning) is used, the amount of calculation required for identification tends to increase particularly. Further, when the camera is installed outdoors, the identification accuracy is likely to change depending on the weather. As an example, when identification is performed based on a captured image in bad weather, the identification accuracy is likely to decrease.

  Therefore, it is desirable to provide a technique that can reduce the amount of calculation required for identification based on the captured image and can perform robust identification against the weather.

  In order to solve the above problem, according to an aspect of the present invention, an image acquisition unit that acquires a captured image captured by a camera, and extraction and reduction of a high-frequency component based on first luminance information of the captured image By performing the above, a recognition device is provided that includes a data generation unit that generates reduced high-frequency information and a discriminator that performs predetermined identification based on the reduced high-frequency information.

  The image acquisition unit may acquire the captured image obtained by capturing the vehicle, and the identifier may identify information related to the vehicle based on the reduced high-frequency information.

  The information related to the vehicle may include at least one of a license plate of the vehicle, a type of the vehicle, a position of the vehicle, and a score of a recognition result based on the reduced high frequency information.

  The discriminator may be a discriminator using a neural network.

  The data generation unit may extract the high frequency component from the first luminance information to generate high frequency information, and reduce the high frequency information to generate the reduced high frequency information.

  The data generation unit may generate the reduced high-frequency information by extracting a representative value for each block from the high-frequency information and arranging the representative value for each block.

  The representative value may be an average value, a maximum value, an intermediate value, or a value of a predetermined pixel position of the block.

  The high frequency component may be a component having a frequency higher than a predetermined threshold.

  The data generation unit may determine the predetermined threshold based on a response value detected from a predetermined target.

  The data generation unit may change the predetermined threshold according to the movement of the object.

  The data generation unit removes the high frequency component from the first luminance information based on the first luminance information and a first filter having a first filter size corresponding to the predetermined threshold. Second luminance information may be generated, and the high frequency information may be generated based on a difference between the first luminance information and the second luminance information.

  The data generation unit generates first reduced luminance information by reducing the first luminance information, and the discriminator generates the first reduced luminance information based on the reduced high frequency information and the first reduced luminance information. Predetermined identification may be performed.

  The data generation unit generates reduced low frequency information by extracting and reducing a frequency component lower than the high frequency component based on the first luminance information, and the discriminator includes the reduced high frequency information and the reduced frequency information. The predetermined identification may be performed based on the reduced low frequency information.

  The data generation unit reduces the second luminance information to generate second reduced luminance information, and generates third reduced luminance information based on the second reduced luminance information and the first filter. And the reduced low-frequency information may be generated based on a difference between the second reduced luminance information and the third reduced luminance information.

  The data generation unit generates third luminance information based on the first luminance information or the second luminance information and a second filter different from the first filter, and the second luminance information Low frequency information may be generated based on a difference between luminance information and the third luminance information, and the reduced low frequency information may be generated by reducing the low frequency information.

  According to another aspect of the present invention, a reduced high frequency signal is obtained by acquiring a captured image captured by a camera and extracting and reducing a high frequency component based on first luminance information of the captured image. A recognition method is provided that includes generating information and performing predetermined identification based on the reduced high-frequency information.

  According to another aspect of the present invention, the computer performs extraction and reduction of a high-frequency component based on an image acquisition unit that acquires a captured image captured by a camera and first luminance information of the captured image. Thus, a program for functioning as a recognition device is provided, which includes a data generation unit that generates reduced high-frequency information and an identifier that performs predetermined identification based on the reduced high-frequency information.

  According to another aspect of the present invention, an image acquisition unit that acquires a captured image captured by a camera, and extraction and reduction of high-frequency components based on first luminance information of the captured image, There is provided a data generation device comprising: a data generation unit that generates reduced high-frequency information and outputs the reduced high-frequency information to a discriminator.

  As described above, according to the present invention, there is provided a technique capable of reducing the amount of calculation required for identification based on a captured image and performing robust identification against the weather.

It is a figure which shows the structural example of the recognition system which concerns on embodiment of this invention. It is a figure which shows the example of luminance information and reduction | decrease information. It is a block diagram which shows the function structural example of the recognition apparatus which concerns on the 1st Embodiment of this invention. It is a figure for demonstrating the example of a filter size. It is a figure for demonstrating the operation example of the data generation part which concerns on the same embodiment. It is a block diagram which shows the function structural example of the recognition apparatus which concerns on the 2nd Embodiment of this invention. It is a figure for demonstrating the operation example of the data generation part which concerns on the same embodiment. It is a block diagram which shows the function structural example of the recognition apparatus which concerns on the 3rd Embodiment of this invention. It is a figure for demonstrating the operation example of the data generation part which concerns on the same embodiment. It is a figure which shows the hardware constitutions of the information processing apparatus as an example of the recognition apparatus which concerns on each embodiment of this invention.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

  In the present specification and drawings, a plurality of constituent elements having substantially the same functional configuration may be distinguished by attaching different numerals to the same reference numerals. However, when there is no need to particularly distinguish each of a plurality of constituent elements having substantially the same functional configuration, only the same reference numerals are given. In addition, similar components in different embodiments may be distinguished by attaching different alphabets after the same reference numerals. However, in the case where it is not necessary to particularly distinguish each of similar components of different embodiments, only the same reference numerals are given.

(0. Overview)
FIG. 1 is a diagram illustrating a configuration example of a recognition system according to an embodiment of the present invention. As shown in FIG. 1, the recognition system 1 according to the embodiment of the present invention includes a recognition device 10 and a camera 20. The camera 20 is provided at a position where the front or back of the vehicle traveling on the road plane P1 can be imaged. The camera 20 is configured to include an image sensor, and images an imaging range by the image sensor. The type of camera 20 is not particularly limited. For example, the camera 20 may be a visible light camera or an infrared light camera.

  FIG. 1 shows an imaging area A1 of the camera 20, and in the imaging area A1, the rear surfaces of the vehicle V1 and the vehicle V2 (running from the left to the right of the page) are imaged by the camera 20. On the other hand, in the imaging region A1, the front surface of the vehicle V3 (running from the right to the left of the page) is imaged by the camera 20. However, the object imaged by the camera 20 is not limited to a vehicle. For example, the object imaged by the camera 20 may be the whole body or part of a person (such as a face), or may be an animal body (such as a vehicle) other than a vehicle.

  The recognition device 10 is connected to the camera 20 and acquires a captured image captured by the camera 20. Then, predetermined discrimination is performed by a discriminator based on the captured image. The type of identification performed by the classifier is not particularly limited as will be described later. However, in the embodiment of the present invention, it is assumed that the camera 20 images a vehicle as an example of the object. Therefore, in the embodiment of the present invention, the classifier identifies information related to the vehicle (for example, a license plate of the vehicle) based on the captured image.

  In recent years, techniques for performing various types of identification based on captured images captured by the camera 20 are known. For example, a technology is disclosed in which the installation position of a license plate is specified based on a captured image and information (for example, a mark, a character string, etc.) described on the license plate is recognized based on the specified installation position. .

  When the identification is performed by the classifier, the luminance information itself of the captured image is input to the classifier. However, when the luminance information itself of the captured image is input to the discriminator, the amount of calculation required for discrimination tends to increase. As an example, when a classifier based on deep learning (deep learning) is used, the amount of calculation required for identification tends to increase particularly. Furthermore, when the camera 20 is installed outdoors, the identification accuracy is likely to change depending on the weather. As an example, when identification is performed based on a captured image in bad weather, the identification accuracy is likely to decrease.

  Therefore, in the embodiment of the present invention, a technique that reduces the amount of calculation required for identification based on a captured image and makes it possible to perform robust identification against the weather will be mainly described. Here, it is also assumed that the luminance information of the captured image is reduced and the reduced luminance information (reduction information) is input to the discriminator.

  FIG. 2 is a diagram illustrating luminance information IMG-11, which is an example of luminance information IMG-luminance information and reduction information, with reference to FIG. 11 shows reduced luminance information IMG-12 obtained by the reduction of 11. As shown in FIG. 2, when the reduced luminance information IMG-12 is input to the discriminator instead of the luminance information IMG-11 itself, it is assumed that the amount of calculation required for the discrimination is reduced. However, when the reduced luminance information IMG-12 is input to the discriminator, the high-frequency component is lost when the luminance information IMG-11 is reduced, so that the identification accuracy is lowered.

  In the embodiment of the present invention, the recognition apparatus 10 extracts a high frequency component from the captured image and performs identification based on the high frequency component. As a result, the amount of calculation required for identification can be reduced, and loss of high-frequency components can be prevented, so that degradation of identification accuracy can also be prevented. Furthermore, since the influence of bad weather (for example, disturbances such as raindrops) is removed by extracting high-frequency components, it becomes possible to perform robust identification with respect to the weather.

  Further, the imaging area A1 may include an area where the camera 20 is in focus (“focus” in FIG. 1) and an area where the camera 20 is not in focus (“defocus” in FIG. 1). . A high-frequency component tends to increase in an in-focus area, and a low-frequency component tends to increase in an out-of-focus area. In the embodiment of the present invention, since the recognition device 10 extracts a high-frequency component from the captured image and performs identification based on the high-frequency component, it is expected that the identification accuracy regarding the vehicle existing in the focused area is improved. Is done.

  The outline of the embodiment of the present invention has been described above.

(1. First embodiment)
Subsequently, a first embodiment of the present invention will be described.

  FIG. 3 is a block diagram illustrating a functional configuration example of the recognition device according to the first embodiment of the present invention. As shown in FIG. 3, the recognition apparatus 10A according to the first embodiment of the present invention includes a generator 100A and a discriminator 160.

  The generator 100A and the identifier 160 include a CPU (Central Processing Unit) and the like, and a function stored in the RAM (Random Access Memory) is executed by the CPU stored in a storage unit (not shown). Can be realized. At this time, a computer-readable recording medium that records the program can also be provided. Alternatively, the generator 100A and the identifier 160 may be configured by dedicated hardware, or may be configured by a combination of a plurality of hardware.

  The generator 100A can function as a data generation device. The generator 100A includes an image acquisition unit 110 and a data generation unit 120A.

  The image acquisition unit 110 acquires a captured image captured by the camera 20. Here, it is assumed that the image acquisition unit 110 acquires a captured image obtained by capturing the vehicle. The data generation unit 120A generates reduced high-frequency information IMG-15 by extracting and reducing high-frequency components based on the luminance information (first luminance information) of the captured image. The discriminator 160 performs predetermined discrimination based on the reduced high-frequency information IMG-15. Here, it is assumed that the identifier 160 identifies information related to the vehicle based on the reduced high-frequency information.

  In the example shown in FIG. 3, it is assumed that the identifier 160 identifies the type of vehicle (vehicle type) as an example of information about the vehicle (the vehicle type is “automobile” or “motorcycle”). The case of identifying whether it is “other” or “other”). However, the information regarding the vehicle is not particularly limited. For example, as an example of information about the vehicle, a license plate of the vehicle (for example, a character string described in the license plate, a position of the license plate, etc.) may be identified, or a position of the vehicle may be identified. The score of the recognition result based on the reduced high-frequency information IMG-15 (such as a score indicating how easy the vehicle type or license plate can be identified) may be identified.

  The classifier 160 may be generated by machine learning, and may be a classifier using a neural network, for example.

  In the example shown in FIG. 3, the discriminator 160 includes a convolution layer 161-1, a pooling layer 161-2, a convolution layer 161-3, a pooling layer 161-4,. 162-1 to 162-N. In other words, the discriminator 160 is configured by a CNN (Convolutional Neural Network). However, the discriminator 160 may be configured by a neural network that does not have a convolution layer and a pooling layer. Further, in the example shown in FIG. 3, the discriminator 160 is configured by a network (deep learning) configured by multi-stage full coupling. However, the total coupling does not have to be configured in multiple stages.

  Hereinafter, data generation by the data generation unit 120A will be mainly described. Specifically, the data generation unit 120A generates high-frequency information by extracting a high-frequency component from the luminance information (first luminance information) of the captured image. Then, the data generation unit 120A generates reduced high-frequency information by reducing high-frequency information. At this time, the high frequency component may be a component having a frequency higher than a predetermined threshold.

  Any method may be used to extract the high-frequency component. For example, an example in which the data generation unit 120A extracts a high-frequency component from luminance information (first luminance information) of a captured image using a DoG (Difference of Gaussian) filter will be mainly described below. However, the data generation unit 120A may extract a high-frequency component from the luminance information (first luminance information) of the captured image by using FFT (fast Fourier transform). When filtering is performed on the luminance information (first luminance information) of the captured image, a filter size corresponding to a predetermined threshold is set.

  FIG. 4 is a diagram for explaining an example of the filter size. Referring to FIG. 4, a license plate N1 is shown, and a character string is written on the license plate N1. The line width W1 indicates the width of characters written on the license plate N1. FIG. 4 shows a filter size W2 (corresponding to one cycle W2 of a kernel function used in the filter). In order to extract the character string written on the license plate N1, it is desirable that the filter size W2 is set to about twice the line width W1 (for example, 2 to 3 times the line width W1). Hereinafter, the filter (first filter) in which the filter size (first filter size) corresponding to the predetermined threshold is set in this way is used.

  FIG. 5 is a diagram for explaining an operation example of the data generation unit 120A according to the first embodiment of the present invention. Referring to FIG. 5, luminance information IMG-11 (first luminance information) of the captured image is shown. Here, the filter (first filter) has a function of reducing sharpness (function of increasing the degree of blur) (that is, a function of removing high-frequency components). Therefore, the data generation unit 120A, based on the luminance information IMG-11 and the filter (first filter), the luminance information IMG-13 (second luminance information) obtained by removing high frequency components from the luminance information IMG-1. Is generated.

  Then, the data generation unit 120A generates the high frequency information IMG-14 based on the difference between the luminance information IMG-11 and the luminance information IMG-13. More specifically, the data generation unit 120A generates the high frequency information IMG-14 by subtracting the luminance information IMG-13 from the luminance information IMG-11 for each pixel. The high frequency information IMG-14 contains only high frequency components. Then, the data generation unit 120A generates reduced high-frequency information IMG-15 by reducing the high-frequency information IMG-14.

  In the reduced high-frequency information IMG-15, the outline of the vehicle remains as an example of the high-frequency component. On the other hand, from the reduced high-frequency information IMG-15, the influence of bad weather (for example, disturbance such as raindrops) is almost removed as an example of other components.

  Here, the method for reducing the high-frequency information IMG-14 is not particularly limited. For example, the data generation unit 120A extracts the representative value for each block from the high frequency information IMG-14, and generates the reduced high frequency information by arranging the representative value for each block. The representative value is not particularly limited. For example, the data generation unit 120A may calculate an average value for each block as an example of the representative value from the high frequency information IMG-14. Alternatively, the data generation unit 120A may extract the maximum value for each block as an example of the representative value from the high frequency information IMG-14.

  Alternatively, the data generation unit 120A may extract an intermediate value for each block as an example of the representative value from the high frequency information IMG-14. Alternatively, the data generation unit 120A may determine the value of a predetermined pixel position of the block (for example, the value of the pixel position at the upper left of the rectangular area of n pixels vertically and horizontally when each block is configured by a rectangular area of n pixels vertically and horizontally) Etc.) may be extracted as an example of the representative value. The reduced high-frequency information IMG-15 generated in this way is input to the discriminator 160.

  In the above description, it is mainly assumed that the data generation unit 120A uses the same threshold value as a threshold value (predetermined threshold value) for extracting high-frequency components. However, the data generation unit 120A may change the threshold value (predetermined threshold value) for extracting the high frequency component in accordance with the movement of the object (vehicle).

  For example, as described with reference to FIG. 1, the high-frequency component tends to increase in the focused area (the central area in the moving direction of the vehicle), and the out-of-focus area (in the moving direction of the vehicle) Low frequency components tend to increase in both sides of the central area. Therefore, the data generation unit 120A may set the threshold used in the in-focus area higher than the threshold used in the out-of-focus area. By doing so, it is expected that the identification accuracy regarding the vehicle is further improved.

  In addition, the data generation unit 120A may determine a threshold (predetermined threshold) for extracting a high-frequency component according to the situation. For example, the data generation unit 120A may determine a predetermined threshold based on a response value detected from a predetermined target. The target is not particularly limited, but may be the license plate of the vehicle traveling on the road plane P1 when the discriminator 160 identifies the license plate of the vehicle.

  For example, the data generation unit 120A generates the high frequency information of the target by the above-described method while changing the threshold value (that is, changing the filter size). Then, the data generation unit 120A may calculate a total result for each pixel of the absolute value of the high-frequency information as a response value, and determine a threshold value that provides the largest response value as a predetermined threshold value used for identification ( The filter size that provides the largest response value may be determined as the filter size used for identification).

  As described above, according to the first embodiment of the present invention, the recognition apparatus 10A extracts a high-frequency component from a captured image and performs identification based on the high-frequency component. As a result, the amount of calculation required for identification can be reduced, and loss of high-frequency components can be prevented, so that degradation of identification accuracy can also be prevented. Furthermore, since the influence of bad weather (for example, disturbances such as raindrops) is removed by extracting high-frequency components, it becomes possible to perform robust identification with respect to the weather.

  The first embodiment of the present invention has been described above.

(2. Second Embodiment)
Subsequently, a second embodiment of the present invention will be described.

  FIG. 6 is a block diagram illustrating a functional configuration example of the recognition apparatus according to the second embodiment of the present invention. As shown in FIG. 6, the recognition device 10 </ b> B according to the second embodiment of the present invention includes a generator 100 </ b> B and a discriminator 160. The generator 100B includes an image acquisition unit 110 and a data generation unit 120B. The recognition apparatus 10B according to the second embodiment of the present invention is mainly different from the recognition apparatus 10A according to the first embodiment of the present invention in that it includes a data generation unit 120B instead of the data generation unit 120A. Therefore, in the following, the data generation unit 120B will be mainly described, and detailed description of other configurations will be omitted.

  In the second embodiment of the present invention, the data generation unit 120B generates the reduced high-frequency information IMG-15 and reduces the luminance information (first luminance information) of the captured image, thereby reducing the reduced luminance information IMG-. 12 (first reduced luminance information) is generated. Then, as illustrated in FIG. 6, the reduced high frequency information IMG-15 and the reduced luminance information IMG-12 are associated (superimposed) with the corresponding pixels of the reduced high frequency information IMG-15 and the reduced luminance information IMG-12 from the data generation unit 120B. Entered. The discriminator 160 performs predetermined discrimination based on the reduced high-frequency information IMG-15 and the reduced luminance information IMG-12.

  FIG. 7 is a diagram for explaining an operation example of the data generation unit 120B according to the second embodiment of the present invention. Referring to FIG. 7, reduced high-frequency information IMG-15 generated by the data generation unit 120B is shown by the same method as that of the first embodiment of the present invention. Furthermore, the data generation unit 120A generates reduced luminance information IMG-12 by reducing the luminance information IMG-11 (first luminance information) of the captured image. The technique for reducing the luminance information IMG-11 may be the same technique as the technique for reducing the high-frequency information IMG-14. The data generation unit 120B outputs the reduced high-frequency information IMG-15 and the reduced luminance information IMG-12 to the discriminator 160, and the discriminator 160 is based on the reduced high-frequency information IMG-15 and the reduced luminance information IMG-12. Predetermined identification is performed.

  As described above, according to the second embodiment of the present invention, as in the first embodiment of the present invention, the amount of calculation required for identification is reduced, and robust identification against weather is performed. It becomes possible. Further, in a scene where identification is possible without color information, such as identification of a license plate, luminance information may not be used for identification, but there may be a scene where color information is necessary for identification. According to the second embodiment of the present invention, it is possible to perform identification with higher accuracy by using luminance information together with high-frequency information for identification. At this time, the data generation unit 120B may determine whether to use the luminance information together with the high frequency information for identification based on the color information of the object.

(3. Third embodiment)
Subsequently, a third embodiment of the present invention will be described.

  FIG. 8 is a block diagram illustrating a functional configuration example of a recognition apparatus according to the third embodiment of the present invention. As shown in FIG. 8, a recognition apparatus 10C according to the third embodiment of the present invention includes a generator 100C and a discriminator 160. The generator 100C includes an image acquisition unit 110 and a data generation unit 120C. The recognition apparatus 10C according to the third embodiment of the present invention is mainly different from the recognition apparatus 10A according to the first embodiment of the present invention in that it includes a data generation unit 120C instead of the data generation unit 120A. Therefore, in the following, the data generation unit 120C will be mainly described, and detailed description of other configurations will be omitted.

  In the third embodiment of the present invention, the data generation unit 120C generates the reduced high-frequency information IMG-15, and has a frequency component lower than the high-frequency component based on the luminance information (first luminance information) of the captured image. The reduced low frequency information IMG-18 is generated by performing extraction and reduction. Then, as shown in FIG. 8, the reduced high frequency information IMG-15 and the reduced low frequency information IMG-18 are associated with each other (superimposed) from the data generation unit 120C, and the discriminator 160 is overlapped. Is input. The discriminator 160 performs predetermined discrimination based on the reduced high frequency information IMG-15 and the reduced low frequency information IMG-18.

  FIG. 9 is a diagram for explaining an operation example of the data generation unit 120C according to the third embodiment of the present invention. Referring to FIG. 9, reduced high-frequency information IMG-15 generated by the data generation unit 120B is shown by the same method as in the first embodiment of the present invention. Furthermore, the data generation unit 120C generates reduced low-frequency information IMG-18 by extracting and reducing a frequency component lower than the high-frequency component based on the luminance information IMG-11 (first luminance information) of the captured image. To do.

  Here, based on the luminance information IMG-13 and a filter (second filter) different from the filter (first filter) used to generate the luminance information IMG-13 by the data generation unit 120C, Assume that the luminance information IMG-16 (third luminance information) is generated. Then, the data generation unit 120C subtracts the luminance information IMG-16 for each pixel from the luminance information IMG-13 based on the difference between the luminance information IMG-13 and the luminance information IMG-16. ), The low frequency information IMG-17 is generated, and the low frequency information IMG-17 is reduced to generate the reduced low frequency information IMG-18. The technique for reducing the low frequency information IMG-17 may be the same technique as the technique for reducing the high frequency information IMG-14.

  Here, the filter may be applied to the luminance information IMG-11 instead of the luminance information IMG-13. That is, the data generation unit 120C may generate the luminance information IMG-16 (third luminance information) based on the luminance information IMG-11 and the filter (second filter). Alternatively, in order to reduce the number of filters to be prepared, the filter (first filter) used for generating the luminance information IMG-13 may be used again.

  That is, the data generation unit 120C may reduce the luminance information IMG-13 first and generate reduced luminance information (second reduced luminance information). Then, the data generation unit 120C, based on the reduced luminance information (second reduced luminance information) and the filter (first filter) used to generate the luminance information IMG-13, the reduced luminance information (third (Reduced luminance information) may be generated. Then, the data generation unit 120C may generate reduced low-frequency information based on the difference between the reduced luminance information (second reduced luminance information) and the reduced luminance information (third reduced luminance information).

  The technique for reducing the luminance information IMG-13 may be the same technique as the technique for reducing the high-frequency information IMG-14. In the reduced high-frequency information IMG-18, almost no outline of the vehicle remains as an example of the high-frequency component. On the other hand, in the reduced high frequency information IMG-18, a line drawn on the road plane is left as an example of a frequency component lower than the high frequency component.

  The data generation unit 120C outputs the reduced high frequency information IMG-15 and the reduced low frequency information IMG-18 to the discriminator 160, and the discriminator 160 converts the reduced high frequency information IMG-15 and the reduced low frequency information IMG-18. Based on the predetermined identification.

  As described above, according to the third embodiment of the present invention, as in the first embodiment of the present invention, the amount of calculation required for identification is reduced, and robust identification against the weather is performed. It becomes possible. Furthermore, the frequency component of the object is not limited to the high frequency component, and may have a frequency component lower than the high frequency component. According to the third embodiment of the present invention, it is possible to perform identification with higher accuracy by using the frequency component lower than the high frequency component together with the high frequency information for identification.

(4. Hardware configuration example)
Next, a hardware configuration example of the recognition device 10 according to each embodiment of the present invention will be described. Below, the hardware structural example of the information processing apparatus 900 is demonstrated as a hardware structural example of the recognition apparatus 10 which concerns on each embodiment of this invention. Note that the hardware configuration example of the information processing apparatus 900 described below is merely an example of the hardware configuration of the recognition apparatus 10. Therefore, as for the hardware configuration of the recognition apparatus 10, an unnecessary configuration may be deleted from the hardware configuration of the information processing apparatus 900 described below, or a new configuration may be added.

  FIG. 10 is a diagram showing a hardware configuration of an information processing apparatus 900 as an example of the recognition apparatus 10 according to each embodiment of the present invention. The information processing apparatus 900 includes a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM (Random Access Memory) 903, a host bus 904, a bridge 905, an external bus 906, an interface 907, , An input device 908, an output device 909, a storage device 910, and a communication device 911.

  The CPU 901 functions as an arithmetic processing device and a control device, and controls the overall operation in the information processing device 900 according to various programs. Further, the CPU 901 may be a microprocessor. The ROM 902 stores programs used by the CPU 901, calculation parameters, and the like. The RAM 903 temporarily stores programs used in the execution of the CPU 901, parameters that change as appropriate during the execution, and the like. These are connected to each other by a host bus 904 including a CPU bus or the like.

  The host bus 904 is connected to an external bus 906 such as a PCI (Peripheral Component Interconnect / Interface) bus via a bridge 905. Note that the host bus 904, the bridge 905, and the external bus 906 are not necessarily configured separately, and these functions may be mounted on one bus.

  The input device 908 includes an input means for inputting information such as a mouse, a keyboard, a touch panel, a button, a microphone, a switch, and a lever, and an input control circuit that generates an input signal based on the input by the user and outputs the input signal to the CPU 901. Etc. A user who operates the information processing apparatus 900 can input various data or instruct a processing operation to the information processing apparatus 900 by operating the input device 908.

  The output device 909 includes, for example, a CRT (Cathode Ray Tube) display device, a liquid crystal display (LCD) device, an OLED (Organic Light Emitting Diode) device, a display device such as a lamp, and an audio output device such as a speaker.

  The storage device 910 is a device for storing data. The storage device 910 may include a storage medium, a recording device that records data on the storage medium, a reading device that reads data from the storage medium, a deletion device that deletes data recorded on the storage medium, and the like. The storage device 910 is configured with, for example, an HDD (Hard Disk Drive). The storage device 910 drives a hard disk and stores programs executed by the CPU 901 and various data.

  The communication device 911 is a communication interface configured with, for example, a communication device for connecting to a network. Further, the communication device 911 may support either wireless communication or wired communication.

  Heretofore, the hardware configuration example of the recognition apparatus 10 according to each embodiment of the present invention has been described.

(5. Summary)
As described above, according to the embodiment of the present invention, the image acquisition unit 110 that acquires a captured image captured by the camera 20 and the extraction and reduction of high-frequency components based on the first luminance information of the captured image. By performing the above, a recognition device 10 is provided that includes a data generation unit 120 that generates reduced high-frequency information and a discriminator 160 that performs predetermined identification based on the reduced high-frequency information.

  According to this configuration, the amount of calculation required for identification can be reduced, and loss of high-frequency components can be prevented, so that degradation of identification accuracy can also be prevented. Furthermore, according to this configuration, the influence of bad weather (for example, disturbance such as raindrops) is removed by extracting the high-frequency component, so that robust identification with respect to the weather can be performed.

  Furthermore, according to the embodiment of the present invention, the image acquisition unit 110 that acquires the captured image captured by the camera 20 and the extraction and reduction of the high-frequency component based on the first luminance information of the captured image, A generator 100 (data generation device) is also provided, which includes a data generation unit 120 that generates reduced high-frequency information and outputs the reduced high-frequency information to the discriminator.

  According to such a configuration, it is possible to generate data for reducing the amount of calculation required for identification. Further, according to such a configuration, it is possible to generate data for performing robust identification with respect to the weather.

  As described above, the program for the recognition device 10 can be provided, but the program for the generator 100 (data generation device) and the program for the discriminator 160 can also be provided separately. At this time, a computer-readable recording medium recording the program of the generator 100 (data generation apparatus) can be provided, and a computer-readable recording medium recording the program of the discriminator 160 can also be provided. .

  The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

  For example, the case where the generator 100B and the discriminator 160 are integrated has been mainly described above. However, the generator 100B and the identifier 160 may be configured separately. In the above description, the case where the camera 20 is configured as a separate body from the recognition device 10 has been mainly described. However, a part or all of the recognition device 10 may be integrated with the camera 20. That is, the camera 20 and the recognition device 10 may be integrated, the camera 20 and the generator 100 may be integrated, or the camera 20 and the identifier 160 may be integrated. Good.

1 recognition system 10 (10A to 10C) recognition device 100 (100A to 100C) generator (data generation device)
DESCRIPTION OF SYMBOLS 110 Image acquisition part 120 (120A-120C) Data generation part 160 Classifier 20 Camera A1 Imaging area P1 Road plane V1-V3 Vehicle

Claims (18)

An image acquisition unit that acquires a captured image captured by the camera;
A data generation unit that generates reduced high-frequency information by extracting and reducing high-frequency components based on the first luminance information of the captured image;
An identifier for performing predetermined identification based on the reduced high-frequency information;
A recognition device comprising: The image acquisition unit acquires the captured image obtained by capturing the vehicle,
The identifier identifies information on the vehicle based on the reduced high-frequency information;
The recognition device according to claim 1. The information on the vehicle includes at least one of a license plate of the vehicle, a type of the vehicle, a position of the vehicle, and a score of a recognition result based on the reduced high frequency information.
The recognition apparatus according to claim 2. The classifier is a classifier using a neural network.
The recognition device according to claim 1. The data generation unit extracts the high frequency component from the first luminance information to generate high frequency information, reduces the high frequency information, and generates the reduced high frequency information.
The recognition device according to claim 1. The data generation unit extracts a representative value for each block from the high-frequency information, and generates the reduced high-frequency information by arranging a representative value for each block.
The recognition apparatus according to claim 5. The representative value is any one of an average value, a maximum value, an intermediate value, and a predetermined pixel position value of the block.
The recognition apparatus according to claim 6. The high frequency component is a component having a frequency higher than a predetermined threshold.
The recognition apparatus according to claim 5. The data generation unit determines the predetermined threshold based on a response value detected from a predetermined target.
The recognition apparatus according to claim 8. The data generation unit changes the predetermined threshold according to the movement of the object.
The recognition apparatus according to claim 8. The data generation unit removes the high-frequency component from the first luminance information based on the first luminance information and a first filter having a first filter size corresponding to the predetermined threshold. Generating second luminance information, and generating the high frequency information based on a difference between the first luminance information and the second luminance information;
The recognition apparatus according to claim 8. The data generation unit generates first reduced luminance information by reducing the first luminance information;
The discriminator performs the predetermined discrimination based on the reduced high-frequency information and the first reduced luminance information.
The recognition device according to claim 1. The data generation unit generates reduced low frequency information by extracting and reducing a frequency component lower than the high frequency component based on the first luminance information,
The discriminator performs the predetermined discrimination based on the reduced high-frequency information and the reduced low-frequency information;
The recognition apparatus according to claim 11. The data generation unit reduces the second luminance information to generate second reduced luminance information, and generates third reduced luminance information based on the second reduced luminance information and the first filter. And generating the reduced low-frequency information based on the difference between the second reduced luminance information and the third reduced luminance information.
The recognition device according to claim 13. The data generation unit generates third luminance information based on the first luminance information or the second luminance information and a second filter different from the first filter, and the second luminance information Generating low frequency information based on a difference between luminance information and the third luminance information, reducing the low frequency information to generate the reduced low frequency information;
The recognition device according to claim 13. Obtaining a captured image captured by the camera;
Generating reduced high-frequency information by extracting and reducing high-frequency components based on the first luminance information of the captured image;
Performing predetermined identification based on the reduced high-frequency information;
A recognition method comprising: Computer
An image acquisition unit that acquires a captured image captured by the camera;
A data generation unit that generates reduced high-frequency information by extracting and reducing high-frequency components based on the first luminance information of the captured image;
An identifier for performing predetermined identification based on the reduced high-frequency information;
A program for functioning as a recognition device. An image acquisition unit that acquires a captured image captured by the camera;
A data generator that generates reduced high-frequency information by extracting and reducing high-frequency components based on the first luminance information of the captured image, and outputs the reduced high-frequency information to a discriminator;
A data generation device comprising:

Images