JP2018206295A - Image recognition device - Google Patents

Abstract

To provide an image recognition device with high recognition accuracy even if there is an installation error in an imaging unit.SOLUTION: The image recognition device comprises: an imaging unit 10 for capturing an image of an imaging object containing the background; a drive mechanism unit 14 for changing an imaging posture of the imaging unit 10; an extraction unit 24 for extracting a target point from a captured image captured by the imaging unit 10; and a processing unit 26 for drive controlling the drive mechanism unit 14 to move the extracted target point to a reference position and executing image processing on the captured image after the movement.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an image recognition apparatus, and more particularly to an image recognition apparatus for camera images.

  As conventional techniques, an imaging unit that images an operation object and a predetermined movable object that can be moved by the action of an external force, and the positions of both objects based on the operation object and the captured image of the movable object by the imaging unit A positive determination result is obtained by the position specifying means for specifying the position, the synchronous movement determining means for determining whether or not the positions of the two objects specified by the position specifying means have moved in synchronization, and the synchronous movement determining means. In such a case, there is known an image recognition apparatus that includes gesture recognition control means that invalidates a gesture recognition result based on a captured image of an operation object and performs gesture recognition (see, for example, Patent Document 1).

  This image recognition apparatus is supposed to be able to reliably extract a recognition object and perform gesture recognition.

JP 2014-6179 A

  However, in the image recognition apparatus of Patent Document 1, the imaging area of the camera that is the imaging means is fixed. For this reason, for example, when the camera is mounted on the vehicle, the recognition area may be shifted due to an assembly error, or the mounting position itself may be changed and the recognition area may be shifted if the mounting model is different. There was a problem that decreased.

  Therefore, an object of the present invention is to provide an image recognition apparatus with high recognition accuracy even when there is an attachment error of an imaging unit.

[1] In order to achieve the above object, an imaging unit that captures an imaging target including a background, a drive mechanism unit that changes the imaging posture of the imaging unit, and a target point are extracted from the captured image captured by the imaging unit And a processing unit that drives and controls the drive mechanism unit, moves the extracted target point to a reference position, and executes image processing on the captured image after the movement. An image recognition apparatus is provided.

[2] The [1] above, wherein the driving mechanism unit controls the imaging point to be imaged at a reference position by changing the imaging posture by a positional movement or a rotational movement of the imaging unit. The image recognition apparatus described in the above may be used.

[3] The image recognition device according to [1] or [2], wherein the target point is a position of a vehicle interior structure of a vehicle.

[4] The image recognition device according to any one of [1] to [3], wherein the target point is set for each vehicle type.

  According to the image recognition device of the present invention, it is possible to provide an image recognition device with high recognition accuracy even if there is an attachment error of the imaging unit.

FIG. 1 is a schematic diagram showing an example of the inside of a vehicle on which an image recognition apparatus according to an embodiment of the present invention is mounted. FIG. 2 is a configuration block diagram showing an example of the configuration of the image recognition apparatus according to the embodiment of the present invention. FIG. 3 is a diagram illustrating movement of the imaging unit in the X, Y, and Z directions and rotational movements α, β, and γ about the X, Y, and Z axes as changes in the imaging posture. FIG. 4A is a diagram illustrating a photographic image of the imaging region captured by the imaging unit, a reference point, and the position of the extracted target point. FIG. 4B is a diagram illustrating the target point as a reference point. FIG. 4C is a diagram in which the target point is moved to the Y position of the reference point, and the target point is made to coincide with the reference point. FIG. 5 is a diagram showing a processing flow of the image processing unit in the image recognition apparatus according to the embodiment of the present invention.

(Embodiment of the present invention)
An image recognition apparatus 1 according to an embodiment of the present invention includes an imaging unit 10 that captures an imaging target including a background, a drive mechanism unit 14 that changes an imaging posture of the imaging unit 10, and an image captured by the imaging unit 10. An extraction unit 24 that extracts a target point from the image, and a processing unit 26 that controls the drive of the drive mechanism unit 14, moves the extracted target point to a reference position, and executes image processing on the captured image after the movement. , And is configured.

  FIG. 1 is a schematic diagram showing an example of the inside of a vehicle on which an image recognition apparatus according to an embodiment of the present invention is mounted. FIG. 2 is a block diagram showing an example of the configuration of the image recognition apparatus according to the embodiment of the present invention. As shown in FIG. 2, the image recognition device 1 according to the embodiment of the present invention includes an imaging unit 10, an image processing unit 20 (an image acquisition unit 22, an extraction unit 24, and a processing unit 26), and a drive circuit 30. It is roughly structured.

(Imaging unit 10)
The imaging unit 10 is an imaging device using a semiconductor element such as a CCD image sensor (Charge Coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor) image sensor. As an example, the imaging unit 10 is disposed in the vicinity of a touch pad 88 disposed on a floor console 80a between a driver seat and a passenger seat of the vehicle 8 as illustrated in FIG. Therefore, the imaging range 15 is an area above the touch pad 88. For example, the imaging unit 10 captures an image of a gesture operation performed by an operator's fingers performed above the touch pad 88. Moreover, this imaging part 10 is comprised so that near infrared rays may be irradiated to the imaging range 15 direction, for example, and it may image.

  As shown in FIG. 2, for example, the imaging unit 10 has a camera element 11 mounted on a camera substrate 12, and a filter 13 is attached to the front surface of the camera element 11. The camera element 11, the camera substrate 12, and the filter 13 are assembled together.

  The imaging unit 10 has a drive mechanism unit 14, and the camera element 11, the camera substrate 12, and the filter 13 can be changed as an integrated fixed body by the drive mechanism unit 14.

  FIG. 3 is a diagram illustrating movement of the imaging unit in the X, Y, and Z directions and rotational movements α, β, and γ about the X, Y, and Z axes as changes in the imaging posture. The camera element 11, the camera substrate 12, and the filter 13 are integrally movable in the X, Y, and Z directions and can be rotated and moved α, β, and γ around the X, Y, and Z axes. . That is, the camera element 11, the camera substrate 12, and the filter 13 are integrally supported by the drive mechanism unit 14 so as to be movable in the X, Y, and Z directions, and are rotated around the X, Y, and Z axes. α, β, and γ are rotatably supported.

  As shown in FIG. 1, for example, the imaging unit 10 is arranged upward on the floor console 80 a with the imaging range 15 facing upward. The arrangement of the imaging unit 10 is not limited to the floor console 80a. For example, the center console 80b or the like may be used so that a gesture made in front of the display device 87 can be captured.

  The imaging unit 10 periodically images the imaging range 15, generates image information S as information of the captured image, and outputs the image information S to the image processing unit 20. As an example, images obtained by imaging the imaging range 15 are a captured image 100 illustrated in FIGS. 4A and 4B and a captured image 110 illustrated in FIG. 4C.

(Drive mechanism 14)
The drive mechanism unit 14 moves in the X, Y, and Z directions shown in FIG. 3 and changes the rotational movements α, β, and γ around the X, Y, and Z axes as shown in FIG. , Each having an actuator (not shown) that enables a roll angle). The actuator is configured to be variable in the X, Y, and Z directions and the rotation directions α, β, and γ around the X, Y, and Z axes. For example, a piezoelectric element can be used. The actuator can drive the drive mechanism unit 14 by changing the voltage to be applied. As the piezo element (electrostrictive element), for example, a laminated type or a bimorph type can be used. The actuator described above is not limited to a piezo element, and can be used as a driving element such as a motor or a solenoid.

(Image processing unit 20)
The image processing unit 20 is roughly configured by an image acquisition unit 22, an extraction unit 24, and a processing unit 26. The image processing unit 20 includes, for example, a CPU (Central Processing Unit) that performs operations and processes on acquired data according to a stored program, a RAM (Random Access Memory) that is a semiconductor memory, a ROM (Read Only Memory), and the like. It is a microcomputer configured. The ROM stores, for example, a program for operating the image processing unit 20. For example, the RAM is used as a storage area for storing a plurality of pieces of image information, temporary calculation results, and the like. In addition, the image processing unit 20 includes a means for generating a clock signal therein, and executes processing based on the clock signal.

(Image acquisition unit 22)
The image acquisition unit 22 acquires the image information S via the camera substrate 12 with respect to the camera element 11 of the imaging unit 10. A captured image 100 is generated from the image information S.

(Extractor 24)
FIG. 4A is a diagram illustrating a photographic image of the imaging region captured by the imaging unit, a reference point, and the position of the extracted target point. FIG. 4B is a diagram illustrating the target point as a reference point. FIG. 4C is a diagram in which the target point is moved to the Y position of the reference point, and the target point is made to coincide with the reference point.

  The extraction unit 24 extracts the target point 210 from the captured image 100 illustrated in FIG. At least one target point is extracted. In the case of a vehicle, the target point can be the vehicle interior structure position of the vehicle 8. For example, specific parts such as a room mirror, a front window part, and a seat belt can be set as target points. Alternatively, a mark (not conspicuous) may be installed in the passenger compartment.

  The target point 210 shown above needs to be included in the background image 300 imaged by the arranged imaging unit 10, and further, an imaging target imaged by the imaging unit 10, for example, an operator who performs a gesture operation It is preferable to set the position where it is difficult to hide with fingers.

  The extraction unit 24 can extract the target point 210 by, for example, a known pattern matching method. For example, the shape of a specific part such as a room mirror, a front window part, and a seat belt is set as a template with a binary image. The extraction unit 24 performs pattern matching using the above-described template using the image of the captured image 100 as a binary image.

As shown in FIG. 4A, the target point 210 is extracted as coordinates (X ₁ , Y ₁ ) by the procedure as described above.

The processing unit 26 changes the imaging posture of the imaging unit 10 and moves the extracted target points (X ₁ , Y ₁ ) to the reference position (X ₀ , Y ₀ ). For example, as illustrated in FIG. 4B, the camera element 11, the camera substrate 12, and the filter 13 are connected via the drive circuit 30 and the drive mechanism unit 14 based on the X-direction difference (X ₁ −X ₀ ). is moved to the reference position X ₀ is driven to integrally X direction. Further, as shown in FIG. 4C, the camera element 11, the camera substrate 12, and the filter 13 are connected via the drive circuit 30 and the drive mechanism unit 14 based on the difference in the Y direction (Y ₁ −Y ₀ ). moving the reference position Y ₀ is driven to integrally Y direction.

The procedure for moving the target points (X ₁ , Y ₁ ) shown above to the reference positions (X ₀ , Y ₀ ) is an example. For example, by combining movement in the X, Y, and Z directions and rotational movement α, β, γ (pitch angle, yaw angle, roll angle), the target point (X ₁ , Y ₁ ) is moved to the reference position (X ₀ , (Y ₀ ) may be moved. Since the movement and rotation are linked to each other, the imaging posture control of the imaging unit 10 is repeatedly performed until the target point (X ₁ , Y ₁ ) converges within a predetermined range of the reference position (X ₀ , Y ₀ ). May be.

By moving the target point (X ₁ , Y ₁ ) to the reference position (X ₀ , Y ₀ ), it is possible to perform image processing for a desired recognition area.

  As illustrated in FIG. 4C, the processing unit 26 can perform image processing on the captured image 110 after the movement. Examples of image processing include binarization processing, shape extraction by pattern matching, and gesture determination based on the extracted shape.

(Operation of image recognition device)
FIG. 5 is a diagram showing a processing flow of the image processing unit in the image recognition apparatus according to the embodiment of the present invention.

  The imaging unit 10 acquires the captured image 100 (step S1). The imaging unit 10 images the imaging range 15, generates image information S as information of the captured image, and outputs the image information S to the image processing unit 20. An image obtained by imaging the imaging range 15 is, for example, a captured image 100 illustrated in FIG.

In the image processing unit 20, the extraction unit 24 extracts a target point (step S2). The extraction unit 24 extracts a target point by, for example, a known pattern matching method. The target point is extracted as (X ₁ , Y ₁ ), for example, as shown in FIG.

Processing unit 26, by changing the imaging position of the imaging unit 10, the extracted target point X _1, is moved to the reference position X ₀ (step S3). For example, as illustrated in FIG. 4B, the camera element 11, the camera substrate 12, and the filter 13 are connected via the drive circuit 30 and the drive mechanism unit 14 based on the X-direction difference (X ₁ −X ₀ ). is moved to the reference position X ₀ is driven to integrally X direction.

Then, the processing unit 26, by changing the imaging position of the imaging unit 10, the extracted target point Y _1, is moved to the reference position Y ₀ (step S4). For example, as shown in FIG. 4C, the camera element 11, the camera substrate 12, and the filter 13 are connected via the drive circuit 30 and the drive mechanism unit 14 based on the difference in the Y direction (Y ₁ −Y ₀ ). moving the reference position Y ₀ is driven to integrally Y direction.

As illustrated in FIG. 4C, the processing unit 26 acquires the captured image 110 after the movement, and executes image processing (step S5). Examples of image processing include binarization processing, shape extraction by pattern matching, and gesture determination based on the extracted shape. These image processes are executed in a state where the target points (X ₁ , Y ₁ ) coincide with the reference positions (X ₀ , Y ₀ ). Therefore, since image processing can be executed within a desired image range, the processing load can be reduced. That is, since the image processing is performed on the captured image 110 that is appropriately limited as a desired image range based on the target point and the reference position from the captured image 110, it is not necessary to perform processing on a useless area, and processing load Can be reduced.

  The operation of the image recognition apparatus is executed as necessary and can be executed repeatedly.

(Effect of embodiment)
The embodiment of the present invention has the following effects.
(1) The image recognition apparatus 1 according to the embodiment of the present invention captures an image with the imaging unit 10 that captures an imaging target including a background, the drive mechanism unit 14 that changes the imaging posture of the imaging unit 10, and the imaging unit 10. The extraction unit 24 that extracts a target point from the captured image and the drive mechanism unit 14 are driven and controlled, the extracted target point is moved to the reference position, and image processing is performed on the moved captured image Part 26. Since the image processing is executed in a state where the target points (X ₁ , Y ₁ ) are coincident with the reference positions (X ₀ , Y ₀ ), the system has a high recognition accuracy even if there is an attachment error. That is, a mechanism capable of changing the angle of the camera board and the XYZ position is provided, and the difference in mounting position and assembly error for each vehicle type can be corrected by the internal mechanism with the position of the vehicle interior structure as a reference point.
(2) When this embodiment is applied to a vehicle, the target point (X ₁ , Y ₁ ) and the reference position (X ₀ , Y ₀ ) can be set corresponding to each vehicle type. Therefore, a single image recognition device 1 (ASSY) can cope with multiple vehicle types, and a structure of a camera image recognition system that can be applied to multiple vehicle types is possible.
(3) The image processing is executed in a state where the target point (X ₁ , Y ₁ ) matches the reference position (X ₀ , Y ₀ ). Therefore, image processing can be executed within a desired image range. Thereby, it is possible to narrow the recognition target area to reduce the amount of calculation, and as a result, there is an effect of speeding up the system.

  As mentioned above, although some embodiment of this invention was described, these embodiment is only an example and does not limit the invention which concerns on a claim. Moreover, these novel embodiments can be implemented in various other forms, and various omissions, replacements, changes, and the like can be made without departing from the scope of the present invention. In addition, not all the combinations of features described in these embodiments are essential to the means for solving the problems of the invention. Furthermore, these embodiments are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 1 ... Image recognition apparatus, 8 ... Vehicle, 10 ... Imaging part, 11 ... Camera element, 12 ... Camera board, 13 ... Filter, 14 ... Drive mechanism part, 15 ... Imaging range, 20 ... Image processing part, 22 ... Image acquisition , 24 ... extraction unit, 26 ... processing unit, 30 ... drive circuit, 80 a ... floor console, 80 b ... center console, 87 ... display device, 88 ... touch pad, 100, 110 ... captured image, 200 ... reference position, 210 ... target point, 300 ... background image

Claims (4)

An imaging unit that captures an imaging target including a background;
A drive mechanism unit for changing an imaging posture of the imaging unit;
An extraction unit for extracting a target point from a captured image captured by the imaging unit;
A drive unit that controls the drive mechanism, moves the extracted target point to a reference position, and executes image processing on the captured image after the movement;
An image recognition apparatus comprising:   2. The image according to claim 1, wherein the driving mechanism unit controls the imaging point to be imaged at a reference position by changing the imaging posture by a positional movement or a rotational movement of the imaging unit. Recognition device.   The image recognition apparatus according to claim 1, wherein the target point is a vehicle interior structure position of the vehicle.   The image recognition apparatus according to claim 1, wherein the target point is set for each vehicle type.