JP2006301031A - Autofocus device and method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve ability of discriminating the direction of a focusing position. <P>SOLUTION: An evaluated value is calculated by using the frequency component of a video signal in a specified area provided in an imaging frame. When discriminating that the evaluated value obtained when wobbling is started is small in ST11, the amplitude of the wobbling is enlarged in ST12. A change in a relative angle is detected in ST13. When the detected change in the relative angle is small, operation progresses from ST15 to ST16, and the amplitude of the wobbling is enlarged. By performing the wobbling in ST17, the direction of the focusing position is discriminated, based on the change in the evaluated value. When the detected change in the relative angle is large, the operation progresses from ST15 to ST17, and the direction of the focusing position is determined without performing the wobbling. The movement of a focus lens in the direction of the focusing position is started, and such a focusing operation that a focal position is moved to the focusing position is performed, based on the evaluated value. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an autofocus device, an autofocus method, and a program. Specifically, the evaluation value is calculated using the frequency component of the video signal in the specific area provided in the imaging image frame, and the direction of the in-focus position is determined based on the change in the evaluation value when the lens is wobbled. In this case, the wobbling amplitude is changed according to the magnitude of the evaluation value at the start of wobbling and the magnitude of the relative angle change between the subject direction and the imaging direction.

  2. Description of the Related Art Conventionally, an imaging apparatus such as a video camera or a digital camera is provided with an autofocus mechanism that automatically focuses on a subject. In this autofocus mechanism, for example, as shown in Patent Document 1, an evaluation value is calculated by adding high-frequency components of a video signal in a specific area provided in a captured image frame, and the focus is set so that the evaluation value becomes maximum. The lens is moved. For this reason, if the specific area is set at the center of the imaging frame and the composition is determined so that the subject is positioned at the center of the imaging frame, the focus position is set to the in-focus position. It can be moved automatically to focus on the subject. Also, when starting the focusing operation to move the focus position to the in-focus position, the wobbling lens is wobbled to prevent the focus position from moving in the direction opposite to the in-focus position. The lens movement direction is determined from the change.

JP-A-10-213736

  By the way, the evaluation value obtained by adding the high frequency components of the video signal is a large value when the contrast of the subject is high, and a small value when the contrast is low. For this reason, even if wobbling is performed when an object with low contrast is imaged, the change in the evaluation value is small, so that it is difficult to determine the direction of the in-focus position.

  For example, as shown in FIG. 13A, when an object with high contrast is being imaged, the change in the evaluation value when the focus position is moved is as shown in FIG. 13B. Further, as shown in FIG. 14A, when an object with low contrast is imaged, the change in the evaluation value when the focus position is moved is as shown in FIG. 14B, compared with when an object with high contrast is imaged. The change in evaluation value is small. Therefore, when wobbling is performed when an object with high contrast is being imaged, the direction of the in-focus position can be determined from the change in the evaluation value. However, when an object with low contrast is being imaged, wobbling is performed. However, since the change of the evaluation value is small, it is difficult to determine the direction of the in-focus position.

  Also, when the subject is moving, when the imaging device is shaking, or when panning is performed in which the imaging direction is swung horizontally, the image in the specific area changes and the evaluation value fluctuates. May end up. In such a case, even if wobbling is performed, it is difficult to determine whether the change in the evaluation value is due to the change in the focus position or the image in the specific area. The direction of the focal position cannot be correctly determined.

  Therefore, the present invention provides an autofocus device, an autofocus method, and a program that can improve the ability to determine the direction of the in-focus position.

  An autofocus device according to the present invention includes a lens driving unit that drives a lens, an evaluation value calculation unit that calculates an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame, and an evaluation value And a control unit that performs a focusing operation to move the focus position to the focusing position based on, and the control unit determines the direction of the focusing position based on a change in the evaluation value when the lens is wobbled, The amplitude of the wobbling is changed according to the evaluation value.

  In addition, another autofocus device according to the present invention includes a lens driving unit that drives a lens, an evaluation value calculation unit that calculates an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame, and A relative angle change detection unit that detects a relative angle change between the subject direction and the imaging direction, and a control unit that performs a focusing operation to move the focus position to the in-focus position based on the evaluation value. The direction of the in-focus position is determined based on the change in the evaluation value when the lens is wobbled, and the amplitude of the wobbling is changed according to the magnitude of the relative angle change detected by the relative angle change detection unit. It is.

  In the autofocus method and program according to the present invention, an evaluation value calculation step for calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame, and an evaluation value when the lens is wobbled The direction of the in-focus position is discriminated based on the change of the in-focus position, the wobbling amplitude is changed in accordance with the relative angle change between the subject direction and the imaging direction, and the in-focus position direction is discriminated. A lens driving processing step for starting a movement of the focus position and performing a focusing operation for moving the focus position to the in-focus position based on the evaluation value, or is executed by a computer.

  In another autofocus method or program according to the present invention, an evaluation value calculation step for calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame, and a lens wobbling A focus position direction determining step for determining the direction of the focus position based on a change in the evaluation value, and changing the wobbling amplitude according to a relative angle change between the subject direction and the imaging direction, and the determined focus A lens drive processing step of starting a movement of the focus position in the direction of the position and performing a focusing operation for moving the focus position to the in-focus position based on the evaluation value, or is executed by a computer.

  In the present invention, for example, an evaluation value indicating a value corresponding to blurring of an image is calculated by adding high-frequency components of a video signal in the evaluation frame provided in the captured image frame. Here, when determining the direction of the in-focus position based on the change in the evaluation value when wobbling the wobbling lens, when the evaluation value at the start of wobbling and the relative angle change between the subject direction and the imaging direction are small, Wobbling is performed by increasing the amplitude of wobbling. This relative angle change is detected using an evaluation value calculated using a luminance component in an evaluation frame provided in the captured image frame.

  According to this invention, when determining the direction of the in-focus position based on the change in the evaluation value when the lens is wobbled, the magnitude of the evaluation value at the start of wobbling and the relative angle change between the subject direction and the imaging direction The amplitude of wobbling is changed in accordance with the size of. For this reason, when the contrast of the subject is low or when shaking occurs during imaging, the ability to determine the in-focus position direction can be improved by changing the wobbling amplitude so that the evaluation value changes significantly. it can.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows the overall configuration of an image pickup apparatus having an autofocus mechanism, for example, a video camera 10.

  The lens block 20 of the video camera 10 includes an imaging lens, a lens position detection unit that detects the position of the imaging lens, a lens driving unit that drives the imaging lens, and the like. In the lens block 20 shown in FIG. 1, as an imaging lens, a focus lens 21 for focusing the subject image on the imaging surface of the imaging device, and a wobbling lens used for determining the direction of the in-focus position. 22 is shown. If the focus lens 21 and the wobbling lens 22 are provided as described above, the focus lens is large, for example, for a professional or professional video camera, and it is difficult to wobble the focus lens in the optical axis direction. Even in this case, it is possible to determine the direction of the in-focus position by wobbling the wobbling lens in the optical axis direction.

  For the focus lens 21, a position detection unit 21a that detects the position of the focus lens 21, that is, a focus position, and a lens drive unit 21b that moves the lens position in the optical axis direction are provided. Similarly, for the wobbling lens 22, a position detection unit 22a and a lens driving unit 22b for moving the lens position in the optical axis direction are provided so that wobbling can be performed correctly. Further, the lens block 20 has an iris 23 for adjusting the amount of light. With respect to the iris 23 as well, an iris position detection unit 23a for detecting the opening state of the iris 23 and an iris drive for opening and closing the iris 23 are provided. A portion 23b is provided.

  The lens block control unit 51 includes a detection signal RSf indicating the focus position from the position detection unit 21a, a detection signal RSw indicating the wobbling amount from the position detection unit 22a, and a detection signal RSi indicating the opening state of the iris from the iris position detection unit 23a. Supplied respectively. The lens block control unit 51 is connected to a user interface 55 for setting an autofocus operation mode and starting an autofocus operation, and an operation signal PSL is sent in response to an operation of the user interface 55. 51. The lens block control unit 51 is provided with a storage unit (not shown) configured using a ROM (or EEPROM) or the like, and the focal length data and aperture ratio data of the focus lens 21 and the wobbling lens 22 are provided. Information such as the manufacturer name and serial number of the lens block is stored.

  The lens block control unit 51 uses the lens based on the stored information, detection signals RSf, RSw, RSi, operation signal PSL, and a focus control signal CTf and a wobbling control signal CTw supplied from a camera block control unit 52 described later. Drive signals RDf and RDw are generated. Further, the generated lens driving signal RDf is supplied to the lens driving unit 21b, and the focus lens 21 is moved so that a desired subject is focused. Further, the generated lens driving signal RDw is supplied to the lens driving unit 22b, and the wobbling lens 22 is wobbled so that the direction of the in-focus position can be detected. Further, the lens block control unit 51 controls the iris opening amount by generating the iris drive signal RDi and supplying it to the iris drive unit 23b.

  The color separation prism 31 of the camera block 30 separates the incident light from the lens block 20 into three primary colors of R (red), G (green), and B (blue), and converts the R component light into an image sensor 32R. The G component light is supplied to the image sensor 32G, and the B component light is supplied to the image sensor 32B.

  The imaging element 32R generates an imaging signal SR corresponding to the R component light by photoelectric conversion and supplies the imaging signal SR to the preamplifier unit 33R. The imaging element 32G generates an imaging signal SG corresponding to the G component light by photoelectric conversion and supplies the imaging signal SG to the preamplifier unit 33G. The imaging element 32B generates an imaging signal SB corresponding to the B component light by photoelectric conversion and supplies the imaging signal SB to the preamplifier unit 33B.

  The preamplifier unit 33R amplifies the level of the imaging signal SR, performs correlated double sampling, removes reset noise, and supplies it to the A / D conversion unit 34R. The A / D converter 34R converts the supplied imaging signal SR into a digital video signal DRa and supplies it to the preprocessing unit 35. The preamplifier unit 33G amplifies the level of the imaging signal SG, performs correlated double sampling, removes reset noise, and supplies the result to the A / D conversion unit 34G. The A / D converter 34G converts the supplied imaging signal SG into a digital video signal DGa and supplies it to the preprocessor 35. The preamplifier unit 33B amplifies the level of the imaging signal SB, performs correlated double sampling, removes reset noise, and supplies it to the A / D conversion unit 34B. The A / D conversion unit 34B converts the supplied imaging signal SB into a digital video signal DBa and supplies it to the preprocessing unit 35.

  The pre-processing unit 35 performs gain adjustment, black level stabilization, dynamic range adjustment, and the like of the supplied video signals DRa, DGa, DBa, and the obtained video signals DRb, DGb, DBb with the signal processing unit 36. The evaluation value calculation unit 37 is supplied.

  The signal processing unit 36 performs various signal processing on the video signals DRb, DGb, and DBb supplied from the preprocessing unit 35 to generate a video output signal DVout. For example, knee correction that compresses a certain level of the video signal, gamma correction that corrects the video signal level according to the set gamma curve, white clip processing that limits the signal level of the video signal to a predetermined range, and black Perform clip processing. The signal processing unit 36 performs contour enhancement processing, linear matrix processing, encoding processing for generating a video output signal DVout in a desired format, and the like.

  The evaluation value calculation unit 37 calculates an evaluation value ID using the frequency component of the video signal, using the video signal in a specific area provided in the captured image frame from the video signals DRb, DGb, and DBb, and performs camera block control. To the unit 52.

  FIG. 2 shows the configuration of the evaluation value calculation unit. The evaluation value calculation unit 37 generates a luminance signal DY based on the video signals DRb, DGb, and DBb, and evaluation values for generating, for example, 14 types of evaluation values ID0 to ID13 as described later. The interface circuit 373 communicates with the generation circuits 372 -ID 0 to 372 -ID 13 and the camera block control unit 52 and supplies the generated evaluation value in response to a request from the camera block control unit 52.

  The luminance signal generation circuit 371 performs an operation (DY = 0.30DRb + 0.59DGb + 0.11DBb) using the video signals DRb, DGb, DBb supplied from the preprocessing unit 35, and generates a luminance signal DY. In this way, the luminance signal DY is generated by determining whether the contrast is high or low in order to determine whether the focus is in focus or not. The change in the contrast indicates the level change of the luminance signal DY. This is because it may be detected.

  The evaluation value generation circuit 372-ID0 generates an evaluation value ID0. Similarly, the evaluation value generation circuits 372-ID1 to 372-ID13 generate evaluation values ID1 to ID13. These evaluation values are basically the sum of the frequency components of the video signal in a specific area (hereinafter referred to as “evaluation frame”) provided in the captured image frame, and indicate values corresponding to image blurring. It is.

Evaluation value ID0: Evaluation value name “IIR1_W1_HPeak”
Evaluation value ID1: Evaluation value name “IIR1_W2_HPeak”
Evaluation value ID2: Evaluation value name “IIR1_W2_HPeak”
Evaluation value ID3: Evaluation value name “IIR4_W3_HPeak”
Evaluation value ID4: Evaluation value name “IIR0_W1_VIntg”
Evaluation value ID5: Evaluation value name “IIR3_W1_VIntg”
Evaluation value ID6: Evaluation value name “IIR1_W1_HIntg”
Evaluation value ID 7: Evaluation value name “Y_W1_HIntg”
Evaluation value ID 8: Evaluation value name “Y_W1_Satul”
Evaluation value ID 9: Evaluation value name “IIR1_W3_HPeak”
Evaluation value ID 10: Evaluation value name “IIR1_W4_HPeak”
Evaluation value ID 11: Evaluation value name “IIR1_W5_HPeak”
Evaluation value ID 12: Evaluation value name “Y_W3_HIntg”
Evaluation value ID 13: Evaluation value name “Y_W3_HIntg”
Here, evaluation value names indicating evaluation value attributes (use data_evaluation frame size_evaluation value calculation method) are assigned to the evaluation values ID0 to ID13.

  The usage data of the evaluation value name is roughly classified into “IIR” and “Y”. There are IIR that uses high-frequency component data extracted from the luminance signal DY using an HPF (high-pass filter), and Y that uses the frequency component of the luminance signal DY as it is without using HPF.

  When HPF is used, IIR type (infinite length impulse response type) HPF is used. Depending on the type of HPF, it is divided into evaluation values IIR0, IIR1, IIR3 and IIR4, which represent HPFs having different cutoff frequencies. In this way, by setting HPFs having different cut-off frequencies, for example, in the vicinity of the in-focus position, using an HPF having a high cut-off frequency is evaluated compared to using an HPF having a low cut-off frequency. The change in value can be increased. Further, when the focus is greatly deviated, the change in evaluation value can be increased by using an HPF having a low cutoff frequency as compared with the case of using an HPF having a high cutoff frequency. In this way, HPFs having different cut-off frequencies are set so that an optimum evaluation value can be selected in accordance with the focus state in the course of the autofocus operation.

  The evaluation frame size is the size of the image area used for generating the evaluation value. As shown in FIG. 3, for example, five types of evaluation frame sizes W1 to W5 are provided, and the center of each evaluation frame coincides with the center of the captured image. FIG. 3 shows the evaluation frame sizes W1 to W5 when the screen size of one field is 768 pixels × 240 pixels.

Evaluation frame size W1: 116 pixels × 60 pixels Evaluation frame size W2: 96 pixels × 60 pixels Evaluation frame size W3: 232 pixels × 120 pixels Evaluation frame size W4: 192 pixels × 120 pixels Evaluation frame size W5: 576 pixels × 180 pixels

  Thus, by setting a plurality of types of frame sizes, different evaluation values corresponding to the respective frame sizes can be generated. Therefore, an appropriate evaluation value can be obtained from any of the evaluation values ID0 to ID13 regardless of the size of the target subject.

  The evaluation value calculation method includes HPeak, HIntg, VIntg, and Satul methods. The HPeak method indicates the peak method horizontal evaluation value calculation method, the HIntg method indicates the total integration method horizontal evaluation value calculation method, the VIntg method indicates the integration method vertical evaluation value calculation method, and the Satul method indicates the number of saturated luminances. ing.

  The HPeak method is an evaluation value calculation method for obtaining a high-frequency component from a horizontal video signal using HPF, and is used to calculate evaluation values ID0, ID1, ID2, ID3, ID9, ID10, and ID11.

  FIG. 4 shows a configuration of a horizontal direction evaluation value calculation filter used in the HPeak method. The horizontal direction evaluation value calculation filter includes an HPF 381 that extracts only a high frequency component from the luminance signal DY of the luminance signal generation circuit, an absolute value processing circuit 382 that takes an absolute value of the high frequency component, and horizontal frame control for the absolute value high frequency component. A multiplication circuit 383 that multiplies the signal WH, a line peak hold circuit 384 that holds one peak value per line, and a vertical direction integration circuit 386 that integrates each peak value in the vertical direction for all the lines in the evaluation frame. is doing.

  A high frequency component is extracted from the luminance signal DY by the HPF 381 and is converted into an absolute value by the absolute value processing circuit 382. Next, the horizontal frame control signal WH is multiplied by the multiplication circuit 383 to obtain an absolute value high-frequency component in the evaluation frame. That is, if the frame control signal WH having a multiplication value “0” outside the evaluation frame is supplied to the multiplication circuit 383, only the absolute value high-frequency component in the horizontal evaluation frame can be supplied to the line peak hold circuit 384. In addition, if the frame control signal WH is set so that the multiplication value becomes smaller in the peripheral portion of the evaluation frame, the edge of the outer frame located in the peripheral portion of the evaluation frame (the bright edge around the evaluation frame) as the focus advances. It is possible to eliminate a noise of the evaluation value due to the influence of intrusion into the frame, a sudden change in the evaluation value due to the shaking of the subject, and the like. The line peak hold circuit 384 holds the peak value for each line.

  The vertical integration circuit 386 adds the held peak value to the evaluation value for each line in the vertical evaluation frame based on the vertical frame control signal WV. This method is called the HPeak method because the peak in the horizontal direction (H) is once held.

  The HIntg method is a method for calculating an evaluation value in the horizontal direction obtained by a total integration method. FIG. 5 shows the configuration of the total integration type horizontal direction evaluation value calculation filter. This filter is the same up to the multiplication circuit 383 as compared with the HPeak horizontal evaluation value calculation filter of FIG. 4, but the horizontal addition circuit 385 converts the absolute value high-frequency component in the evaluation frame in the horizontal direction. All are added, and thereafter the vertical integration circuit 386 integrates the addition results of all the vertical lines in the evaluation frame in the vertical direction.

  This total integration type horizontal direction evaluation value calculation filter is used to calculate evaluation values ID6, ID7, ID12, and ID13. Compared with the HPeak method, the HPeak method obtains one peak value per line and adds them in the vertical direction, whereas the HIntg method has an absolute value within the horizontal evaluation frame of each line. The difference is that all the high frequency components are added and added in the vertical direction.

  In the HIntg system, usage data is classified into IIR1 using high-frequency components and Y using the luminance signal DY itself as it is. The luminance addition value is obtained by a luminance addition value calculation filter circuit obtained by removing the HPF 381 from the full integration type horizontal direction evaluation value calculation filter of FIG.

  The VIntg method is a total evaluation type vertical direction evaluation value calculation method, and is used for the evaluation values ID4 and ID5. Both the HPeak method and the HIntg method generate an evaluation value by adding in the horizontal direction, while the VIntg method is an evaluation value generated by adding a high frequency component in the vertical direction. For example, if the upper half of the screen is white and the lower half is black, or a horizontal line, only the high-frequency component in the vertical direction is present and there is no high-frequency component in the horizontal direction, the HPeak method horizontal evaluation value does not function effectively. Therefore, the evaluation value of the VIntg method is determined so that autofocus functions effectively in such a scene.

  FIG. 6 shows a configuration of a vertical direction evaluation value calculation filter for calculating a vertical direction evaluation value. The vertical direction evaluation value calculation filter includes a horizontal direction average value calculation filter 391, an IIR type HPF 392, an absolute value processing circuit 393, and an integration circuit 394.

  The horizontal direction average value calculation filter 391 selects the luminance signal of the pixels (for example, 64 pixels) in the center of the horizontal evaluation frame based on the frame control signal WHc from the luminance signal DY of each line, and calculates an average value (total) The same applies to the value) and is output once per horizontal period. Here, the reason why the 64 pixels are in the center is to remove noise around the evaluation frame. Here, since only 64 pixels are sequentially stored and one average value is finally output, the memory device such as a line memory or a frame memory is not required. Next, this is synchronized with the line frequency, the high frequency component is extracted by the HPF 392, and the absolute value processing circuit 393 converts it into the absolute value high frequency component. Further, the integration circuit 394 integrates all the lines in the vertical evaluation frame based on the vertical frame control signal WV.

  The Satul method is a calculation method for obtaining the number of saturated luminance signals DY within the evaluation frame (specifically, the luminance level is a predetermined amount or more), and is used for calculating the evaluation value ID8. In the calculation of the evaluation value ID8, the luminance signal DY is compared with the threshold value α, and the number of pixels having the luminance signal DY equal to or larger than the threshold value α is counted for each field to evaluate the evaluation value ID8. To do.

  In the reference signal generation unit 40, a vertical synchronization signal VD, a horizontal synchronization signal HD, and a reference signal CLK that are the reference of the operation of each unit in the video camera 10 are generated and supplied to the image sensor driving unit 42. The image sensor driving unit 42 generates a drive signal RIR based on the supplied vertical synchronization signal VD, horizontal synchronization signal HD, and reference signal CLK, supplies the drive signal RIR to the image sensor 32R, and drives the image sensor 32R. Similarly, drive signals RIG and RIB are generated and supplied to the image sensors 32G and 32B to drive the image sensors 32G and 32B. Note that the preamplifier units 33R, 33G, and 33B, the A / D conversion units 34R, 34G, and 34B, the preprocessing unit 35, the signal processing unit 36, the evaluation value calculation unit 37, and the like are vertically synchronized with the video signal supplied from the preceding stage. Processing is performed using the synchronization signal VD, the horizontal synchronization signal HD, and the reference signal CLK. These signals may be supplied from the reference signal generation unit 40 or may be supplied from the previous stage together with the video signal.

  A user interface 56 is connected to the camera block control unit 52, and a control signal is generated based on the operation signal PSC and the like supplied from the user interface 56 and is supplied to each unit. Is controlled based on the operation signal PSC or the like.

  Further, the lens block control unit 51 and the camera block control unit 52 described above can communicate with each other using a predetermined format, protocol, or the like. The lens block control unit 51 and the camera block control unit 52 perform auto communication. Controls the focus operation.

  Here, as described above, for example, the lens block control unit 51 supplies various information (for example, a focus position and an iris value) QF to the camera block control unit 52 in response to a request. Further, the lens drive signals RDf and RDw are generated based on the focus control signal CTf and the wobbling control signal CTw supplied from the camera block control unit 52, and the focus lens 21 and the wobbling lens 22 are driven. Based on the evaluation value ID calculated by the evaluation value calculation unit 37 and various information read from the lens block control unit 51, the camera block control unit 52 performs a focus control signal CTf for driving and controlling the focus lens 21, wobbling, and the like. A wobbling control signal CTw for driving and controlling the lens 22 is generated and supplied to the lens block control unit 51.

  The lens block control unit 51 and the camera block control unit 52 may be configured integrally. In the following description, the lens block control unit 51 and the camera block control unit 52 are collectively referred to as the control unit 50. To do. The control unit 50 may be configured using a microcomputer, a memory, and the like, and may perform an autofocus operation by reading and executing a program stored in the memory.

  Next, the autofocus operation of the video camera 10 will be described. FIG. 7 shows a flowchart of the autofocus operation.

  In step ST1, the control unit 50 starts calculating the evaluation value. In step ST <b> 2, the control unit 50 performs a focus position direction determination process, and performs lens drive setting so that the moving direction of the lens becomes the focus position. In this focus position direction determination process, the lens is wobbled, and the focus position direction is determined based on the change in the evaluation value at this time. As shown in FIG. 1, when the focus lens 21 and the wobbling lens 22 are provided, the wobbling lens is wobbled. When the wobbling lens 22 is not provided, the focus lens is wobbled.

  In addition, when imaging a low-contrast subject and performing wobbling, the control unit 50 increases the change in the evaluation value so that the direction can be easily determined. To change the wobbling amplitude. Furthermore, even if a relative angle change between the subject direction and the imaging direction (for example, a change in the angle of the subject direction with respect to the imaging direction) occurs, the relative angle change so that the direction can be easily determined from the change in the evaluation value. The wobbling amplitude is changed according to the above.

  FIG. 8 is a flowchart showing the focus position direction determination process. In step ST11, the control unit 50 determines whether or not the evaluation value ID is small. Here, when the contrast of the subject is low and the evaluation value ID is equal to or less than the threshold value, the process proceeds to step ST12. Further, since the contrast of the subject is not low, the process proceeds to step ST13 when the evaluation value ID is not less than or equal to the threshold value.

  The evaluation value ID used in step ST11 is calculated using the frequency component of the video signal in the specific area provided in the captured image frame, and the value changes according to the contrast of the subject. For example, the above-described evaluation value ID0 and evaluation value ID6 are used.

  In step ST12, the control unit 50 performs amplitude enlargement setting, performs setting so that the wobbling amplitude is enlarged from the standard state, and proceeds to step ST17. As for the amplitude setting, the setting of the standard state and the setting of expanding the amplitude are stored in advance, and when the evaluation value ID is equal to or less than the threshold value, the setting of expanding the amplitude is read and reflected in the wobbling operation. The amplitude enlargement ratio may be changed according to the evaluation value so that the amplitude increases as the evaluation value decreases. Thus, if the magnification rate of the amplitude is changed according to the evaluation value, the amplitude is not so large when the contrast is not so low, so that it is possible to prevent the time required for the wobbling operation from becoming long.

  When the process proceeds from step ST11 to step ST13, the control unit 50 detects a relative angle change. In this relative angle change detection, a relative angle change between the subject and the imaging direction, that is, a subject shake on the imaging screen caused by a motion of the subject or a shake of the video camera 10 is detected.

  The shake of the subject can be detected using an evaluation value calculated using a luminance component in a specific area provided in the captured image frame. Here, for example, a luminance integrated value is used as the evaluation value calculated using the luminance component, and the shake of the subject can be detected from the normalized difference value of the evaluation value ID7 as in the conventional case. If the evaluation value ID0 is not small, the shake of the subject can be detected using the evaluation value ID0.

  The normalized difference value [%] of the evaluation value ID0 is 50 × | e0 [0] −e2 [0] | / using the current field value e0 [0] and the previous field e2 [0]. It is defined by e0 [0]. Similarly, the normalized difference value [%] of the evaluation value ID7 is 50 × | e0 [7] −e2 [using the value e0 [7] in the current field and the value e2 [7] two fields before that. 7] | / e0 [7].

  The normalized difference value means the rate of change of the evaluation value per field. Here, by comparing the evaluation value of the current field with the evaluation value of the previous two fields, the comparison is performed only in either the odd or even field. Therefore, the evaluation value fluctuates due to the difference between the odd field and the even field. Can be removed. In addition, the reason why it is multiplied by 50 regardless of the percentage is due to the same reason.

  If a moving average value of three fields is used as e0 [0], e2 [0], e0 [7], e2 [7], for example, an indoor fluorescent lamp is driven with a commercial power of 50 Hz. When the video camera 10 is operated at 60 Hz, this influence can be eliminated even if flickering of the fluorescent lamp occurs at 20 Hz.

  Here, the normalized difference value is calculated using the evaluation value ID0 over 12 fields in the shake determination period, and the maximum value among the normalized difference values is set as the maximum normalized difference value (hereinafter, “ndiff_e [0”). )) And shake is judged.

  However, when the value of the evaluation value ID0 is low, “ndiff_e [0]” may be a large value even if the subject is not shaken due to noise fluctuation that exists constantly, and may exceed the shake determination threshold.

  Therefore, when the value of the evaluation value ID0 is low, each normalized difference value of 12 fields is calculated using the evaluation value ID7, and the maximum normalized difference value “ndiff_e [7]” is created from the maximum value. , Shake is judged using it. In all cases, the maximum normalized difference value “ndiff_e [0]” of the evaluation value ID0 is not used if the evaluation value ID0 is equal to or less than a certain threshold value. It is because it thinks that it is doing.

  In the shake determination, for example, the following criteria are used. If the average value of the evaluation value ID0 (average value of the shake determination period 12 field) is 200 or more, the still mode when ndiff_e [0] <3%, the shake mode 1 when ndiff_e [0] ≧ 3%, and ndiff_e [0 ] ≧ 30%, the shaking mode 2 in which the shaking is larger than the shaking mode 1 is set. Further, when the average value of the evaluation value ID0 is less than 200 and ndiff_e [7] <7%, the still mode, when ndiff_e [7] ≧ 7%, the shaking mode 1, and ndiff_e [7] ≧ 12.5% In some cases, shake mode 2 is set. The video camera 10 is provided with an angular velocity sensor, and a sensor signal is supplied from the angular velocity sensor to the control unit 50. As described above, if the angular velocity sensor is provided, the shake of the video camera 10 can be determined based on the sensor signal.

  In step ST14, it is determined whether or not a relative angle change is detected. Here, when it is determined that the mode is the still mode, it is determined that there is no relative angle change, and the process proceeds to step ST17. That is, when there is no change in the relative angle, there is no change in the evaluation value due to the change in the relative angle, so there is no need to increase the wobbling amplitude. If it is determined that the shaking mode 1 or the shaking mode 2 is detected, the process proceeds to step ST15 on the assumption that a relative angle change has been detected.

  In step ST15, the control unit 50 determines whether or not the relative angle change is large. Here, when the relative angle change is not large, that is, when it is determined that the shaking mode is small, for example, the shaking mode 1, the process proceeds to step ST16. On the other hand, when the relative angle change is large, that is, the shake is large, for example, the shake mode 2 is determined, the process proceeds to step ST18.

  In step ST16, similarly to step ST12, the control unit 50 performs amplitude expansion setting and proceeds to step ST17.

  In step ST17, the control unit 50 performs a wobbling process to wobble the wobbling lens 22 in the optical axis direction, and determines the direction of the in-focus position from the change in the evaluation value at this time.

  9 to 11 show the wobbling processing operation. 9 to 11 show a case where the wobbling lens 22 is moved to the Far side after moving to the Near side, but may be moved to the Near side after moving to the Far side.

  The control unit 50 stores the above-described 14 types of evaluation values as the evaluation value em0st [i] (where i = 0 to 13) at time t0. Next, the wobbling lens 22 is moved to the Near side, for example, and each evaluation value when the amplitude is maximum is stored as an evaluation value em1ne [i] at time t1. The time from the time point t0 to the time point t1 is obtained by adding the moving time of the wobbling lens 22 and the phase delay time (one field) generated when the evaluation value calculation unit 37 obtains the evaluation value.

  Next, the wobbling lens 22 is moved to the Far side, and each evaluation value when the amplitude is maximum is stored as an evaluation value em2fa [i] at time t2. Thereafter, the wobbling lens 22 is moved in the direction of the initial position, and each evaluation value when the wobbling lens 22 returns to the initial position is stored as an evaluation value em3nu [i] at time t3. Note that the time from time t1 to time t2 and the time from time t2 to time t3 are also the movement time of the wobbling lens 22 and the phase delay time (1 field) generated when the evaluation value calculation unit 37 obtains the evaluation value. Is added.

  Using the evaluation values em0st [i], em1ne [i], em2fa [i], and em3nu [i] stored in this way, the same direction determination as in the prior art is performed. For example, a calculation process using the evaluation value difference “em0st [i] −em1ne [i]” and the evaluation value difference “em3nu [i] −em2fa [i]” is performed, and whether or not the focus position is based on the calculation result It is assumed that the lens is not moved when the in-focus position is determined. When the in-focus position is not determined, it is based on the calculation result using the difference between the evaluation value em2fa [i] and the evaluation value em1ne [i], and the direction in which the evaluation value increases is the Near direction or Far. The direction of movement of the lens is determined by determining whether the direction is the direction. Further, in the direction determination, first, the first stage determination is performed using the evaluation value of the usage data “IIR1” having a relatively high cutoff frequency to determine the direction. If the direction is not known, the direction determination is performed. The evaluation value that contributes to is changed, and the second stage determination is continued using the evaluation value of the usage data “IIR4” having a comparatively low cutoff frequency.

  FIG. 9 shows a case where the contrast of the subject is not low. When the wobbling process is performed, the evaluation value changes greatly as the wobbling lens 22 moves. FIG. 10 shows a case where the contrast is low. When the contrast is low, the amplitude expansion setting is performed and the wobbling amplitude is increased. Therefore, the change in the evaluation value also increases. In FIG. 10, a broken line indicates a case where the amplitude expansion setting is not performed. FIG. 11 shows a case where the relative angle change is small. When the relative angle change is small, the amplitude expansion setting is performed and the wobbling amplitude is increased. Therefore, the change in the evaluation value also increases. In FIG. 11, a broken line indicates a case where the amplitude expansion setting is not performed.

  In this way, when the evaluation value is small because of low contrast, or when the relative angle change is small with little shaking or subject movement, the change in the evaluation value becomes significant by increasing the amplitude of wobbling, and the focus position is changed. The direction discrimination ability can be improved.

  By the way, when the shaking is large, as shown in FIG. 12, it is difficult to determine whether the change in the evaluation value is due to the fluctuation of the focus position or the shaking even if the wobbling amplitude is increased. Therefore, when the process proceeds from step ST15 to step ST18, the control unit 50 determines the direction of the in-focus position without performing wobbling.

  In the imaging lens, the distance to the subject and the focus position are not proportional. For example, when the focus position is moved on the near end side, the distance difference between the focused subject positions is small. Further, when the focus position is moved on the far end side, the distance difference between the subject positions that are in focus is large. In general, the subject is often several meters or more away, and there are few opportunities to image the approaching subject. Accordingly, since the focus position is often on the Far end side rather than the Near end side, the direction of the focus position is set to the Far direction without performing wobbling.

  In the direction determination, the direction of the in-focus position may be determined based on the current focus position. For example, when the current focus position is close to the Far end, the direction of the focus position is the Near direction.

  As described above, after determining the direction of the in-focus position by the in-focus position / direction determination process in step ST2, the control unit 50 performs the lens driving process in step ST3. That is, the control unit 50 moves the focus lens 21 in the direction of the in-focus position, and performs the same hill-climbing control process as the conventional one.

  In the hill-climbing control process, the increase / decrease in the evaluation value calculated by the evaluation value calculation unit 37 is detected, and the focus position FPs is moved so as to reach the in-focus position FPj so that the evaluation value becomes maximum. In the hill-climbing control process using this evaluation value, for example, the focus lens 21 is moved so that the above-described evaluation values ID0, ID2, etc. are maximized. Further, by using the evaluation value ID8, when the number of pixels with high luminance increases, the evaluation frame size W1 is switched to the evaluation frame size W5 so that the focus lens 21 is not moved in the direction in which blur occurs. The evaluation value is calculated. Furthermore, by using the evaluation value ID0 and other evaluation values ID1 to ID7 and ID9 to ID13, it is possible to perform switching of the lens moving speed, determination of shaking, determination of reverse feed, determination of the near end of the lens and the arrival of the Far end, and the like. And the driving operation of the focus lens 21 is controlled so that focusing is performed with high accuracy based on the determination result.

  In step ST4, the control unit 50 determines whether or not the focus position direction determination process needs to be performed again. For example, when the focus lens 21 is moved, a high-luminance edge enters the evaluation value frame and the evaluation value increases (so-called false mountain of the evaluation value), and lens drive control corresponding to the entry of the high-luminance edge is performed. This may occur when the subject cannot be focused. For this reason, false mountain determination is performed, and when it is determined to be false mountain, the process returns to step ST2, and the focus position direction determination process is performed again. If it is not determined to be a false mountain, the process proceeds to step ST5.

  In step ST5, the control unit 50 determines whether or not to complete the lens driving process. Here, the control unit 50 determines whether or not the focus position FPs has been moved to the in-focus position FPj by the hill-climbing control process, and determines that the focus position FPs has been moved to the in-focus position FPj. If not, the process returns to step ST3. If it is determined that the focus position FPs has been moved to the in-focus position FPj, the lens driving process is completed and the autofocus operation ends.

  In this way, when determining the direction of the in-focus position based on the change in the evaluation value when the wobbling lens is wobbled, the wobbling amplitude is adaptive, and the magnitude of the evaluation value at the start of wobbling and the subject direction By changing the amplitude of the wobbling according to the magnitude of the relative angle change with respect to the imaging direction, the ability to discriminate the direction of the in-focus position can be improved. For example, even when the change in evaluation value due to wobbling cannot be discriminated, such as when the contrast of the subject is low or when shaking occurs during imaging, the wobbling amplitude is increased and the change in evaluation value due to wobbling is significant. Therefore, it is possible to determine the direction of the in-focus position by changing the evaluation value, and it is possible to improve the ability to determine the direction of the in-focus position.

  Further, when the relative angle change is larger than the set threshold value, that is, when the movement of the subject or the shake at the time of imaging is large, the direction of the in-focus position is determined without performing wobbling. For this reason, it is possible to prevent the direction of the in-focus position from being erroneously determined due to a change in the evaluation value due to the movement of the subject or the shaking during imaging. Further, when the relative angle change is larger than the set threshold, wobbling is not performed, so that the lens driving process can be started promptly. In addition, since the relative angle change is detected by using an evaluation value calculated using a luminance component in a specific area provided in the captured image frame, the direction of the in-focus position is determined only by using a video signal. be able to.

  If an angular velocity sensor is provided, the sensor signal from the angular velocity sensor can be used to accurately detect shaking during imaging. By using this sensor signal, the wobbling amplitude can be changed or wobbling can be performed. Processing that is not performed can also be performed with high accuracy.

  In the above-described embodiment, the case where the imaging device is a video camera has been described. However, the present invention can be similarly applied to an imaging device such as a digital camera.

It is a figure which shows the structure of a video camera. It is a figure which shows the structure of an evaluation value calculation part. It is a figure which shows the evaluation frame size. It is a figure which shows the structure of a horizontal direction evaluation value calculation filter. It is a figure which shows the structure of a total integration system horizontal direction evaluation value calculation filter. It is a figure which shows the structure of a vertical direction evaluation value calculation filter. It is a flowchart which shows an autofocus operation | movement. It is a flowchart which shows a focus position direction discrimination | determination process. It is a figure for demonstrating the wobbling operation | movement when contrast is high. It is a figure for demonstrating the wobbling operation | movement when contrast is low. It is a figure for demonstrating the wobbling operation | movement when a relative angle change is small. It is a figure for demonstrating the wobbling operation | movement when a relative angle change is large. It is a figure for demonstrating the change of the evaluation value when contrast is high. It is a figure for demonstrating the change of the evaluation value when contrast is low

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Video camera, 20 ... Lens block, 21 ... Focus lens, 21a, 22a ... Position detection part, 21b, 22b ... Lens drive part, 22 ... Wobbling lens, 23. ..Iris, 23a ... Iris position detector, 23b ... Iris drive, 30 ... Camera block, 31 ... Color separation prism, 32R, 32B, 32B ... Image sensor, 33R, 33G , 33B: Preamplifier unit, 34R, 34G, 34B ... A / D conversion unit, 35 ... Pre-processing unit, 36 ... Signal processing unit, 37 ... Evaluation value calculation unit, 40 ... Reference signal generation unit, 42... Image sensor drive unit, 51... Lens block control unit, 52... Camera block control unit, 55 and 56. Generation circuit, 372-ID0 to 372-ID13 ... evaluation value generation circuit, 373 ... interface circuit, 381,392 ... high-pass filter, 382,393 ... absolute value processing circuit, 383 ... Multiplier circuit, 384 ... line peak hold circuit, 385 ... horizontal direction adder circuit, 386 ... vertical direction integrator circuit, 391 ... horizontal direction average value calculation filter, 394 ... integrator circuit

Claims (9)

A lens driving unit for driving the lens;
An evaluation value calculation unit that calculates an evaluation value using the frequency component of the video signal in a specific area provided in the captured image frame;
A control unit that performs a focusing operation to move the focus position to the focusing position based on the evaluation value;
The control unit determines a direction of the in-focus position based on a change in the evaluation value when the lens is wobbled, and changes the wobbling amplitude according to the magnitude of the evaluation value. Autofocus device. A lens driving unit for driving the lens;
An evaluation value calculation unit that calculates an evaluation value using the frequency component of the video signal in a specific area provided in the captured image frame;
A relative angle change detection unit that detects a relative angle change between the subject direction and the imaging direction;
A control unit that performs a focusing operation to move the focus position to the focusing position based on the evaluation value;
The control unit determines a direction of the in-focus position based on a change in the evaluation value when the lens is wobbled, and determines the amplitude of the wobbling of the relative angle change detected by the relative angle change detection unit. An autofocus device characterized by being changed according to the size. The evaluation value calculation unit calculates an evaluation value using a luminance component in a specific area provided in the captured image frame,
The autofocus device according to claim 2, wherein the relative angle change detection unit detects the relative angle change using an evaluation value calculated using the luminance component. Provide an angular velocity sensor to detect changes in the imaging direction,
The autofocus device according to claim 3, wherein the control unit changes the amplitude of the wobbling in accordance with a detection result of the angular velocity sensor. 3. The autofocus device according to claim 2, wherein the control unit determines a direction of the in-focus position without performing the wobbling when a relative angle change is larger than a set threshold value. 4. An evaluation value calculation step of calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame;
In-focus position direction determination that determines the direction of the in-focus position based on the change in the evaluation value when the lens is wobbled, and the amplitude of the wobbling changes according to the magnitude of the evaluation value at the start of the wobbling Process,
A lens drive processing step of performing a focusing operation of starting the movement of the focus position in the direction of the determined focus position and moving the focus position to the focus position based on the evaluation value. Autofocus method. An evaluation value calculation step of calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame;
A focus position direction determination step of determining a direction of a focus position based on a change in the evaluation value when the lens is wobbled, and changing an amplitude of the wobbling in accordance with a relative angle change between the subject direction and the imaging direction. When,
A lens drive processing step of performing a focusing operation of starting the movement of the focus position in the direction of the determined focus position and moving the focus position to the focus position based on the evaluation value. Autofocus method. On the computer,
An evaluation value calculation step of calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame;
In-focus position direction determination that determines the direction of the in-focus position based on the change in the evaluation value when the lens is wobbled, and the amplitude of the wobbling changes according to the magnitude of the evaluation value at the start of the wobbling Process,
A program for executing a lens drive processing step of starting a movement of the focus position in the direction of the determined focus position and performing a focus operation for moving the focus position to the focus position based on the evaluation value. On the computer,
An evaluation value calculation step of calculating an evaluation value using a frequency component of a video signal in a specific area provided in a captured image frame;
A focus position direction determination step of determining a direction of a focus position based on a change in the evaluation value when the lens is wobbled, and changing an amplitude of the wobbling in accordance with a relative angle change between the subject direction and the imaging direction. When,
A program for executing a lens drive processing step of starting a movement of the focus position in the direction of the determined focus position and performing a focus operation for moving the focus position to the focus position based on the evaluation value.