WO2015008663A1 - Focus control device, and focus control method - Google Patents

Focus control device, and focus control method Download PDF

Info

Publication number
WO2015008663A1
WO2015008663A1 PCT/JP2014/068200 JP2014068200W WO2015008663A1 WO 2015008663 A1 WO2015008663 A1 WO 2015008663A1 JP 2014068200 W JP2014068200 W JP 2014068200W WO 2015008663 A1 WO2015008663 A1 WO 2015008663A1
Authority
WO
WIPO (PCT)
Prior art keywords
filter
focus
unit
value
absolute value
Prior art date
Application number
PCT/JP2014/068200
Other languages
French (fr)
Japanese (ja)
Inventor
周一 朝原
Original Assignee
株式会社Ｊｖｃケンウッド
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社Ｊｖｃケンウッド filed Critical 株式会社Ｊｖｃケンウッド
Publication of WO2015008663A1 publication Critical patent/WO2015008663A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B7/00—Mountings, adjusting means, or light-tight connections, for optical elements
    • G02B7/28—Systems for automatic generation of focusing signals
    • G02B7/36—Systems for automatic generation of focusing signals using image sharpness techniques, e.g. image processing techniques for generating autofocus signals
    • G02B7/365—Systems for automatic generation of focusing signals using image sharpness techniques, e.g. image processing techniques for generating autofocus signals by analysis of the spatial frequency components of the image
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/67—Focus control based on electronic image sensor signals
    • H04N23/673—Focus control based on electronic image sensor signals based on contrast or high frequency components of image signals, e.g. hill climbing method

Definitions

  • the present disclosure relates to a focus control device and a focus control method.
  • the contrast-type focus control apparatus obtains the optimum focus position as follows.
  • the focus control device integrates the contrast values extracted from the entire or partial area of the captured image frame to generate a contrast evaluation value.
  • the focus control apparatus moves the position of the focus lens little by little and sets the position where the contrast evaluation value is maximized as the optimum focus position.
  • a contrast evaluation value is generated by extracting a contrast value, which is a high-frequency component, from a captured image using a high-pass filter (HPF) and integrating the absolute values of the extracted contrast values.
  • a contrast value which is a high-frequency component
  • the performance of focus control depends on the characteristics of HPF.
  • the cut-off frequency of the HPF is low, an edge component with a low frequency can be extracted, so that the focus detection sensitivity becomes high.
  • low-frequency noise is also taken out, it is not suitable for a noisy image.
  • Patent Literature 1 describes that a plurality of HPFs having different characteristics are prepared and the HPFs are adaptively switched depending on the state of an image.
  • Patent Document 2 describes that the influence of random noise is avoided by creating a vector map of contrast values, selecting only portions where vectors of adjacent pixels are aligned to some extent, and integrating the contrast values. Yes.
  • Patent Document 1 has a complicated algorithm for switching HPF.
  • Patent Document 2 requires a large amount of hardware for creating a vector map.
  • An object of the present embodiment is to provide a focus control device and a focus control method capable of focus control with a relatively simple configuration, high focus detection sensitivity, and less susceptible to noise.
  • an image of a certain region including an imaging unit having a focus lens, and a pixel of interest in the video signal output from the imaging unit and peripheral pixels located around the pixel of interest.
  • a plurality of filters having a plurality of filter characteristics which are set to detect a correlation with a texture component equal to or lower than a predetermined level corresponding to the noise level and have different sensitivity levels for detecting the texture component.
  • a filter operation unit that performs a filter operation and obtains a plurality of correlation detection values of the texture component included in the image of the certain region and each of the plurality of filter characteristics, and an absolute value that converts each of the plurality of correlation detection values into an absolute value
  • a binarizing unit an adder for adding the plurality of correlation detection values absolute-valued by the absolute-value converting unit, and an output from the adder.
  • An integration unit that integrates the added value within a predetermined area, and a lens control unit that controls the position of the focus lens in accordance with the focus evaluation value, using the integration value output from the integration unit as a focus evaluation value.
  • a predetermined region corresponding to a noise level is applied to an image in a certain region including a target pixel in the video signal output from the imaging unit and peripheral pixels located around the target pixel.
  • the constant is calculated by using a plurality of filters having a plurality of filter characteristics that are set so as to detect a correlation with a texture component below the level and have different sensitivity levels for detecting the texture component.
  • a plurality of correlation detection values of the texture component of the image of the region and each of the plurality of filter characteristics are obtained, and each of the plurality of correlation detection values is converted into an absolute value to generate a plurality of absolute value signals, and the plurality of absolute values
  • the signals are added to generate an added signal, and the added signal is integrated within a predetermined area to generate an integrated value.
  • the integrated value is used as a focus evaluation value, and the Focus control method characterized by controlling the position of the focus lens where the image pickup unit is provided in accordance with the carcass evaluation value is provided.
  • the focus control device and the focus control method of the embodiment it is possible to perform focus control with a relatively simple configuration, high focus detection sensitivity, and less susceptible to noise.
  • FIG. 1 is a block diagram illustrating an imaging apparatus including the focus control apparatus of the first embodiment.
  • FIG. 2 is a diagram illustrating an example of the filter coefficient of the two-dimensional HPF 11 in FIG.
  • FIG. 3 is a diagram illustrating an example of the filter coefficient of the two-dimensional HPF 12 in FIG.
  • FIG. 4 is a diagram illustrating an example of input / output characteristics of the mixture ratio calculation unit 26 in FIG.
  • FIG. 5 is a block diagram illustrating an imaging apparatus including the focus control apparatus of the second embodiment.
  • FIG. 6 is a diagram illustrating an example of the filter characteristics of the frequency domain filter 30 in FIG.
  • FIG. 7 is a diagram illustrating an example of input / output characteristics of the mixture ratio calculation unit 32 in FIG. 5.
  • the focus control device of each embodiment has a configuration in which the video noise reduction device described in Patent Document 3 is developed into a focus control device.
  • An imaging apparatus 100 illustrated in FIG. 1 includes an imaging unit 10, two-dimensional high-pass filters (two-dimensional HPF) 11 and 12, absolute value conversion units 13 and 14, an adder 15, an integration unit 16, and a lens control unit 17. Further, the imaging apparatus 100 includes a low-pass filter (LPF) 21, a delay circuit unit 22, multipliers 23 and 24, an adder 25, and a mixture ratio calculation unit 26.
  • LPF low-pass filter
  • the two-dimensional HPFs 11 and 12, the absolute value conversion units 13 and 14, the adder 15, the LPF 21, the delay circuit unit 22, the multipliers 23 and 24, the adder 25, and the mixing ratio calculation unit 26 constitute a video noise reduction device.
  • the imaging unit 10, the two-dimensional HPFs 11 and 12, the absolute value converting units 13 and 14, the adder 15, the integrating unit 16, and the lens control unit 17 constitute a focus control device.
  • the imaging unit 10 includes an optical system including a focus lens 10FL and the like, and an imaging signal processing unit that generates a video signal by photoelectrically converting a subject imaged on an imaging element (not shown) through the optical system. (Not shown).
  • the imaging unit 10 supplies the generated video signal to the two-dimensional HPFs 11 and 12, the LPF 21, and the delay circuit unit 22.
  • the video signal includes various video noises.
  • the LPF 21 averages the signals of a plurality of pixels in a certain area including the target pixel in the input video signal and the peripheral pixels located around the target pixel, and the video noise component is equal to or lower than a preset cutoff frequency. Outputs a low-frequency signal with suppressed.
  • the delay circuit unit 22 delays the input video signal by the same time as the delay time generated by the LPF 21, and outputs a delayed video signal time-synchronized with the low frequency signal output from the LPF 21.
  • FIG. 2 shows an example of the filter coefficient of the two-dimensional HPF 11.
  • FIG. 3 shows an example of the filter coefficient of the two-dimensional HPF 12.
  • the two-dimensional HPFs 11 and 12 have coefficients set for each of the 25 pixels, with a central pixel in a 25-pixel planar area of 5 horizontal pixels and 5 vertical pixels as a target pixel. .
  • the two-dimensional HPF 11 has a filter characteristic that is highly sensitive to a texture pattern in a vertical or horizontal direction and low in sensitivity to a texture pattern in an oblique direction.
  • the two-dimensional HPF 11 may have a filter characteristic that is highly sensitive to both the vertical and horizontal texture patterns.
  • the two-dimensional HPF 12 has a filter characteristic that is highly sensitive to a texture pattern in an oblique direction and low in sensitivity to a texture pattern in a vertical or horizontal direction.
  • the filter coefficients shown in FIGS. 2 and 3 are composed of only three kinds of integers of 1, 0, and ⁇ 1 in order to reduce the amount of calculation, and the sum is zero. Therefore, when a 25-pixel image centered on the target pixel of the video signal has a uniformly flat luminance value and only video noise is superimposed, the correlation detection value detected by the two-dimensional HPFs 11 and 12 is 0 or almost 0.
  • the two-dimensional HPFs 11 and 12 have different sensitivity levels for detecting a texture pattern as described above with respect to an image of 25 pixels centered on a target pixel in an input video signal.
  • a filter operation using a filter having filter characteristics is performed.
  • the two-dimensional HPFs 11 and 12 output numerical correlation detection results as a result of the filter operation.
  • the two-dimensional HPF 11 has a filter characteristic in which the luminance difference between the vertical and horizontal images is equal to or lower than a predetermined level corresponding to the noise level, and the sensitivity is high with respect to the vertical or horizontal texture pattern. Therefore, the two-dimensional HPF 11 outputs the correlation detection result having the largest value when the texture component included in the input 25-pixel image matches the texture pattern in the vertical or horizontal direction equivalent to the noise level.
  • the two-dimensional HPF 12 has a filter characteristic in which the luminance difference of the image in the oblique direction is not more than a predetermined level corresponding to the noise level and the sensitivity to the texture pattern in the oblique direction is high. Therefore, the two-dimensional HPF 12 outputs the correlation detection result having the largest value when the texture component included in the input 25-pixel image matches the texture pattern in the oblique direction equivalent to the noise level.
  • the two-dimensional HPFs 11 and 12 are filter arithmetic units that perform a filter operation on an image in a certain region including the target pixel and peripheral pixels located around the target pixel.
  • the two-dimensional HPFs 11 and 12 are set so as to detect a correlation with a texture component equal to or lower than a predetermined level corresponding to the noise level, and have a plurality of filter characteristics having different sensitivity levels for detecting the texture component. Filter calculation is performed using a plurality of filters.
  • the two-dimensional HPFs 11 and 12 obtain a plurality of correlation detection values between a texture component included in an image in a certain region and each of a plurality of filter characteristics by a filter operation.
  • the absolute value converting unit 13 converts the correlation detection value output from the two-dimensional HPF 11 into an absolute value, and supplies the first absolute value signal to the adder 15.
  • the absolute value converting unit 14 converts the correlation detection value output from the two-dimensional HPF 12 into an absolute value, and supplies the second absolute value signal to the adder 15.
  • the adder 15 adds the first absolute value signal and the second absolute value signal, and supplies the addition signal to the mixing ratio calculation unit 26.
  • the adder 15 does not simply add the first absolute value signal and the second absolute value signal, but the larger absolute value of the first absolute value signal and the second absolute value signal. It is good also as a structure which selects a signal.
  • the mixing ratio calculation unit 26 outputs a mixing ratio ⁇ corresponding to the addition value indicated by the addition signal output from the adder 15.
  • FIG. 4 is an example of input / output characteristics of the mixing ratio calculator 26, where the horizontal axis indicates the value of the input signal (ie, the added value) and the vertical axis indicates the value of the output signal (ie, the mixing ratio ⁇ ). As shown in FIG. 4, the mixture ratio calculation unit 26 outputs 0 as the mixture ratio ⁇ in the range from 0 to the threshold value TH11.
  • the mixing ratio calculation unit 26 outputs 1 as the mixing ratio ⁇ in the range where the addition value is equal to or greater than the threshold value TH12.
  • the mixing ratio calculator 26 outputs a value that linearly increases the range from 0 to 1 as the mixing ratio ⁇ in the range where the addition value is from the threshold value TH11 to the threshold value TH12.
  • the thresholds TH11 and TH12 are appropriately set based on experiments or the like so that a desired noise reduction effect can be obtained.
  • the mixing ratio calculation unit 26 supplies the mixing ratio ⁇ calculated according to the input / output characteristics shown in FIG.
  • the multiplier 23 multiplies the low frequency signal output from the LPF 21 by (1- ⁇ ) and outputs the result.
  • the multiplier 24 multiplies ⁇ of the delayed video signal output from the delay circuit unit 22 and outputs the result.
  • the adder 25 adds the output signals of the multipliers 23 and 24 and outputs the result.
  • the mixing ratio ⁇ is 1, an image of 25 pixels can be regarded as an image mainly composed of texture components.
  • the output signal of the multiplier 23 is 0. Therefore, the video noise reduction device outputs the delayed video signal output from the delay circuit unit 22 as it is. If the video noise reduction device outputs the delayed video signal as it is, it means that the resolution is prioritized.
  • the mixing ratio ⁇ When the mixing ratio ⁇ is 0, an image of 25 pixels can be regarded as an image in which video noise is superimposed on a portion having a flat luminance value.
  • the output signal of the multiplier 24 When the mixing ratio ⁇ is zero, the output signal of the multiplier 24 is zero. Therefore, the video noise reduction device outputs the low frequency signal output from the LPF 21 as it is. Outputting the low-frequency signal output from the LPF 21 as it is means giving priority to the noise reduction effect.
  • the video noise reduction device mixes the low frequency signal output from the LPF 21 and the delayed video signal output from the delay circuit unit 22 in accordance with the value of the mixing ratio ⁇ when the mixing ratio ⁇ is in the range from 0 to 1. Output mixed signal.
  • the video noise reduction device can reduce the video noise contained in the flat portion of the image while suppressing the image quality deterioration of the texture portion of the video signal without increasing the hardware scale.
  • the video noise reduction device can reduce video noise even for video signals for which encoding information cannot be obtained.
  • the video noise reduction device can significantly suppress pseudo-noise even for a video signal including a texture component only in a specific direction.
  • the integration unit 16 integrates the addition signal output from the adder 15 for each pixel for a predetermined area.
  • the predetermined area is the whole of one frame or a partial area within one frame.
  • the integrating unit 16 supplies the integrated value to the lens control unit 17 as a focus evaluation value.
  • the lens control unit 17 controls the position of the focus lens 10FL included in the imaging unit 10 to be the in-focus position based on the input focus evaluation value.
  • the lens control unit 17 moves the focus lens 10FL little by little within the lens movable range, and monitors the magnitude of the focus evaluation value.
  • the lens control unit 17 determines the position of the focus lens 10FL when the focus evaluation value is maximized as the focus position.
  • the lens control unit 17 controls the focus lens 10FL so that the position of the focus lens 10FL is always the in-focus position.
  • the two-dimensional HPF 11 for generating the focus control value has high sensitivity to the vertical or horizontal texture pattern and low sensitivity to the diagonal texture pattern.
  • the two-dimensional HPF 12 for generating the focus control value has high sensitivity to the texture in the oblique direction and low sensitivity to the texture pattern in the vertical or horizontal direction.
  • the two-dimensional HPFs 11 and 12 are extremely low in sensitivity to random noise.
  • the filter coefficients of the two-dimensional HPFs 11 and 12 are composed of only ⁇ 1, 0, and 1 so that the calculation amount is small, and the sum is zero.
  • the focus control device of this embodiment has a relatively simple configuration and high focus detection sensitivity. According to the focus control apparatus of the present embodiment, autofocus that is less susceptible to noise is possible.
  • the focus control device of this embodiment is configured to share the two-dimensional HPF 11, the two-dimensional HPF 12, the absolute value converting unit 13, the absolute value converting unit 14, and the adder 15 with the video noise reducing device. Therefore, the cost of the focus control device can be greatly reduced.
  • the two-dimensional HPF 11, the two-dimensional HPF 12, the absolute value conversion unit 13, the absolute value conversion unit 14, and the adder 15 may not be shared with the video noise reduction device but may be provided as a configuration dedicated to the focus control device. Even with such a configuration, there is an effect that auto-focusing is possible with a relatively simple configuration, high focus detection sensitivity, and hardly affected by noise.
  • the imaging apparatus 200 includes a frequency domain filter 30, an absolute value conversion unit 31, a mixture ratio calculation unit 32, multipliers 33 and 34, and an adder 35 in addition to the configuration included in the imaging apparatus 100.
  • the imaging unit 10 supplies the generated video signal to the frequency domain filter 30.
  • the frequency domain filter 30 is composed of a band pass filter (BPF) or HPF that extracts a predetermined frequency domain.
  • BPF band pass filter
  • HPF HPF that extracts a predetermined frequency domain.
  • FIG. 6 is an example of the filter characteristics that the frequency domain filter 30 has.
  • the absolute value conversion unit 31 absoluteizes the output of the frequency domain filter 30.
  • the output of the frequency domain filter 30 has a small value when the contrast is low, and a large value when the contrast is high.
  • the absolute value signal output from the absolute value converting unit 31 is supplied to the mixture ratio calculating unit 32 and the multiplier 34.
  • the mixing ratio calculation unit 32 compares the input signal (the absolute value signal from the absolute value conversion unit 31) with the threshold values TH21 and TH22. As shown in FIG. 7, the mixture ratio calculation unit 32 outputs 0 as the mixture ratio ⁇ in the range where the absolute value signal is from 0 to the threshold value TH21. The mixing ratio calculator 26 outputs 1 as the mixing ratio ⁇ in the range where the absolute value signal is equal to or greater than the threshold value TH21.
  • the mixing ratio calculation unit 32 outputs a value that linearly increases the range from 0 to 1 as the mixing ratio ⁇ when the absolute value signal is in the range from the threshold value TH21 to the threshold value TH22.
  • the mixing ratio calculation unit 32 supplies the mixing ratio ⁇ calculated according to the input / output characteristics shown in FIG.
  • the multiplier 33 multiplies the added signal output from the adder 15 by (1- ⁇ ) and outputs the result.
  • the multiplier 34 multiplies the absolute value signal output from the absolute value converting unit 31 by ⁇ and outputs the result.
  • the adder 35 adds the output signals of the multipliers 33 and 34 and outputs the result.
  • the integration unit 16 integrates the addition signal output for each pixel from the adder 35 for a predetermined area, and supplies the integration value to the lens control unit 17 as a focus evaluation value.
  • the lens control unit 17 controls the position of the focus lens 10FL included in the imaging unit 10 to be the in-focus position based on the input focus evaluation value.
  • the frequency domain filter 30 that extracts a predetermined frequency domain has low contrast and the image includes the frequency component of the pass domain of the frequency domain filter 30, it may be difficult to separate the video noise from the texture component.
  • the outputs of the two-dimensional HPFs 11 and 12 that are correlation detection filters are used when the contrast is low, and the output of the frequency domain filter 30 is used when the contrast is high, or the mixing ratio between the two is changed. It is configured to change. This suppresses the influence of the video noise on the output of the frequency domain filter 30 when the contrast is low.
  • the second embodiment while taking advantage of the characteristics of the frequency domain filter 30, it is possible to suppress the influence of video noise at the time of low contrast and improve the performance of focus control.
  • the filter calculation unit may be configured using three or more two-dimensional HPFs having different sensitivity levels.
  • the present invention includes a computer program for realizing the functions of the respective components of the focus control device in FIGS. 1 and 5 by a computer.
  • the computer program may be taken into the computer from a recording medium, or distributed via a network and downloaded to the computer.
  • the present invention can be used for a contrast type focus control device.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Optics & Photonics (AREA)
  • Automatic Focus Adjustment (AREA)
  • Studio Devices (AREA)
  • Focusing (AREA)

Abstract

The present invention comprises an imaging unit (10) which has a focus lens (10FL). Filter operation units (11, 12) are set to detect, in a video signal, a correlation between a fixed region image and a texture component that is equal to or less than a prescribed level corresponding to a noise level, said fixed region image containing a pixel of interest and peripheral pixels located at the periphery of the pixel of interest. The filter operation units (11, 12) perform a filter operation by using a plurality of filters having a plurality of filter characteristics differing from one another in the directions of high and low sensitivity for detecting the texture component, and determine a plurality of correlation detection values of the texture component that pertains to the fixed region image, and of the respective plurality of filter characteristics. Absolute value conversion units (13, 14) convert each of the plurality of correlation detection values into absolute values. An adder (15) adds the plurality of correlation detection values which have been converted into absolute values. An integration unit (16) integrates the added value in a prescribed region. A lens control unit (17) sets the integrated value as a focus evaluation value, and controls the position of the focus lens (10FL) in accordance with the focus evaluation value.

Description

フォーカス制御装置及びフォーカス制御方法Focus control device and focus control method
　本開示は、フォーカス制御装置及びフォーカス制御方法に関する。 The present disclosure relates to a focus control device and a focus control method.
　カメラのフォーカス制御方式の１つにコントラスト方式と称されるものがある。コントラスト方式のフォーカス制御装置は、次のようにして最適フォーカス位置を求める。 One of the camera focus control methods is called the contrast method. The contrast-type focus control apparatus obtains the optimum focus position as follows.
　フォーカス制御装置は、撮影画像のフレームの全体または部分的な領域から抽出したコントラスト値を積算して、コントラスト評価値を生成する。フォーカス制御装置は、フォーカスレンズの位置を少しずつ動かして、コントラスト評価値が最大になる位置を、最適フォーカス位置とする。 The focus control device integrates the contrast values extracted from the entire or partial area of the captured image frame to generate a contrast evaluation value. The focus control apparatus moves the position of the focus lens little by little and sets the position where the contrast evaluation value is maximized as the optimum focus position.
　一般的に、コントラスト評価値は、撮影画像からハイパスフィルタ（ＨＰＦ）を用いて高域成分であるコントラスト値を抽出し、抽出したコントラスト値の絶対値を積算することによって生成される。 Generally, a contrast evaluation value is generated by extracting a contrast value, which is a high-frequency component, from a captured image using a high-pass filter (HPF) and integrating the absolute values of the extracted contrast values.
　フォーカス制御（オートフォーカス）の性能は、ＨＰＦの特性に左右される。ＨＰＦのカットオフ周波数が低いと、低い周波数のエッジ成分を取り出せるため、合焦の検出感度が高くなる。しかしながら、低い周波数のノイズも取り出されてしまうため、ノイズの多い画像には適さない。 The performance of focus control (autofocus) depends on the characteristics of HPF. When the cut-off frequency of the HPF is low, an edge component with a low frequency can be extracted, so that the focus detection sensitivity becomes high. However, since low-frequency noise is also taken out, it is not suitable for a noisy image.
　ＨＰＦのカットオフ周波数を高くすると、低い周波数のノイズの影響は受けにくくなるものの、エッジ成分も高い周波数のみになるため、合焦の検出感度が低くなる。 When the cutoff frequency of the HPF is increased, the influence of low frequency noise is less likely to be affected, but the edge component also has only a high frequency, so the focus detection sensitivity is lowered.
　この問題点を解決するための手法として、特許文献１及び特許文献２に記載の技術がある。特許文献１には、特性の異なる複数のＨＰＦを用意し、画像の状態によってＨＰＦを適応的に切り替えることが記載されている。特許文献２には、コントラスト値のベクトルマップを作成し、隣接する画素のベクトルがある程度そろっている部分だけを選択してコントラスト値を積算することにより、ランダムノイズの影響を避けることが記載されている。 There are techniques described in Patent Literature 1 and Patent Literature 2 as a technique for solving this problem. Patent Document 1 describes that a plurality of HPFs having different characteristics are prepared and the HPFs are adaptively switched depending on the state of an image. Patent Document 2 describes that the influence of random noise is avoided by creating a vector map of contrast values, selecting only portions where vectors of adjacent pixels are aligned to some extent, and integrating the contrast values. Yes.
特許第４６７９１７９号公報Japanese Patent No. 4679179 特開２００４－０２９１０５号公報JP 2004-029105 A 特開２０１３－１６８８３７号公報JP 2013-168837 A
　しかしながら、特許文献１記載の技術は、ＨＰＦを切り替えるアルゴリズムが複雑である。特許文献２記載の技術は、ベクトルマップを作成するためにハードウェアが大きくなってしまう。 However, the technique described in Patent Document 1 has a complicated algorithm for switching HPF. The technique described in Patent Document 2 requires a large amount of hardware for creating a vector map.
　本実施形態は、比較的簡単な構成で、合焦の検出感度が高く、かつ、ノイズの影響を受けにくいフォーカス制御が可能なフォーカス制御装置及びフォーカス制御方法を提供することを目的とする。 An object of the present embodiment is to provide a focus control device and a focus control method capable of focus control with a relatively simple configuration, high focus detection sensitivity, and less susceptible to noise.
　実施形態の第１の態様によれば、フォーカスレンズを有する撮像部と、前記撮像部から出力された映像信号における注目画素と前記注目画素の周辺に位置する周辺画素とを含む一定領域の画像に対し、ノイズレベルに相当する所定のレベル以下のテクスチャ成分との相関を検出するように設定され、テクスチャ成分を検出するための感度の高低の方向が互いに異なる複数のフィルタ特性を有する複数のフィルタを用いてフィルタ演算して、前記一定領域の画像が有するテクスチャ成分と前記複数のフィルタ特性それぞれとの複数の相関検出値を求めるフィルタ演算部と、前記複数の相関検出値それぞれを絶対値化する絶対値化部と、前記絶対値化部によって絶対値化された前記複数の相関検出値を加算する加算器と、前記加算器より出力された加算値を所定の領域内で積算する積算部と、前記積算部より出力された積算値をフォーカス評価値とし、前記フォーカス評価値に応じて前記フォーカスレンズの位置を制御するレンズ制御部とを備えることを特徴とするフォーカス制御装置が提供される。 According to the first aspect of the embodiment, an image of a certain region including an imaging unit having a focus lens, and a pixel of interest in the video signal output from the imaging unit and peripheral pixels located around the pixel of interest. On the other hand, a plurality of filters having a plurality of filter characteristics which are set to detect a correlation with a texture component equal to or lower than a predetermined level corresponding to the noise level and have different sensitivity levels for detecting the texture component. A filter operation unit that performs a filter operation and obtains a plurality of correlation detection values of the texture component included in the image of the certain region and each of the plurality of filter characteristics, and an absolute value that converts each of the plurality of correlation detection values into an absolute value A binarizing unit, an adder for adding the plurality of correlation detection values absolute-valued by the absolute-value converting unit, and an output from the adder. An integration unit that integrates the added value within a predetermined area, and a lens control unit that controls the position of the focus lens in accordance with the focus evaluation value, using the integration value output from the integration unit as a focus evaluation value. There is provided a focus control device characterized by comprising:
　実施形態の第２の態様によれば、撮像部から出力された映像信号における注目画素と前記注目画素の周辺に位置する周辺画素とを含む一定領域の画像に対し、ノイズレベルに相当する所定のレベル以下のテクスチャ成分との相関を検出するように設定され、テクスチャ成分を検出するための感度の高低の方向が互いに異なる複数のフィルタ特性を有する複数のフィルタを用いてフィルタ演算して、前記一定領域の画像が有するテクスチャ成分と前記複数のフィルタ特性それぞれとの複数の相関検出値を求め、前記複数の相関検出値それぞれを絶対値化して複数の絶対値信号を生成し、前記複数の絶対値信号を加算して加算信号を生成し、前記加算信号を所定の領域内で積算して積算値を生成し、前記積算値をフォーカス評価値とし、前記フォーカス評価値に応じて前記撮像部が備えるフォーカスレンズの位置を制御することを特徴とするフォーカス制御方法が提供される。 According to the second aspect of the embodiment, a predetermined region corresponding to a noise level is applied to an image in a certain region including a target pixel in the video signal output from the imaging unit and peripheral pixels located around the target pixel. The constant is calculated by using a plurality of filters having a plurality of filter characteristics that are set so as to detect a correlation with a texture component below the level and have different sensitivity levels for detecting the texture component. A plurality of correlation detection values of the texture component of the image of the region and each of the plurality of filter characteristics are obtained, and each of the plurality of correlation detection values is converted into an absolute value to generate a plurality of absolute value signals, and the plurality of absolute values The signals are added to generate an added signal, and the added signal is integrated within a predetermined area to generate an integrated value. The integrated value is used as a focus evaluation value, and the Focus control method characterized by controlling the position of the focus lens where the image pickup unit is provided in accordance with the carcass evaluation value is provided.
　実施形態のフォーカス制御装置及びフォーカス制御方法によれば、比較的簡単な構成で、合焦の検出感度が高く、かつ、ノイズの影響を受けにくいフォーカス制御が可能となる。 According to the focus control device and the focus control method of the embodiment, it is possible to perform focus control with a relatively simple configuration, high focus detection sensitivity, and less susceptible to noise.
図１は、第１実施形態のフォーカス制御装置を含む撮像装置を示すブロック図である。FIG. 1 is a block diagram illustrating an imaging apparatus including the focus control apparatus of the first embodiment. 図２は、図１中の２次元ＨＰＦ１１のフィルタ係数の一例を示す図である。FIG. 2 is a diagram illustrating an example of the filter coefficient of the two-dimensional HPF 11 in FIG. 図３は、図１中の２次元ＨＰＦ１２のフィルタ係数の一例を示す図である。FIG. 3 is a diagram illustrating an example of the filter coefficient of the two-dimensional HPF 12 in FIG. 図４は、図１中の混合比計算部２６の入出力特性の一例を示す図である。FIG. 4 is a diagram illustrating an example of input / output characteristics of the mixture ratio calculation unit 26 in FIG. 図５は、第２実施形態のフォーカス制御装置を含む撮像装置を示すブロック図である。FIG. 5 is a block diagram illustrating an imaging apparatus including the focus control apparatus of the second embodiment. 図６は、図５中の周波数領域フィルタ３０が有するフィルタ特性の一例を示す図である。FIG. 6 is a diagram illustrating an example of the filter characteristics of the frequency domain filter 30 in FIG. 図７は、図５中の混合比計算部３２の入出力特性の一例を示す図である。FIG. 7 is a diagram illustrating an example of input / output characteristics of the mixture ratio calculation unit 32 in FIG. 5.
　以下、各実施形態のフォーカス制御装置を、添付図面を参照しながら説明する。各実施形態のフォーカス制御装置は、特許文献３に記載された映像ノイズ低減装置をフォーカス制御装置に発展させた構成を有する。 Hereinafter, the focus control device of each embodiment will be described with reference to the accompanying drawings. The focus control device of each embodiment has a configuration in which the video noise reduction device described in Patent Document 3 is developed into a focus control device.
＜第１実施形態＞
　図１に示す撮像装置１００は、撮像部１０、２次元ハイパスフィルタ（２次元ＨＰＦ）１１，１２、絶対値化部１３，１４、加算器１５、積算部１６、レンズ制御部１７を備える。また、撮像装置１００は、ローパスフィルタ（ＬＰＦ）２１、遅延回路部２２、乗算器２３，２４、加算器２５、混合比計算部２６を備える。 <First Embodiment>
An imaging apparatus 100 illustrated in FIG. 1 includes an imaging unit 10, two-dimensional high-pass filters (two-dimensional HPF) 11 and 12, absolute value conversion units 13 and 14, an adder 15, an integration unit 16, and a lens control unit 17. Further, the imaging apparatus 100 includes a low-pass filter (LPF) 21, a delay circuit unit 22, multipliers 23 and 24, an adder 25, and a mixture ratio calculation unit 26.
　２次元ＨＰＦ１１，１２、絶対値化部１３，１４、加算器１５、ＬＰＦ２１、遅延回路部２２、乗算器２３，２４、加算器２５、混合比計算部２６は、映像ノイズ低減装置を構成する。 The two- dimensional HPFs 11 and 12, the absolute value conversion units 13 and 14, the adder 15, the LPF 21, the delay circuit unit 22, the multipliers 23 and 24, the adder 25, and the mixing ratio calculation unit 26 constitute a video noise reduction device.
　撮像部１０、２次元ＨＰＦ１１，１２、絶対値化部１３，１４、加算器１５、積算部１６、レンズ制御部１７は、フォーカス制御装置を構成する。 The imaging unit 10, the two- dimensional HPFs 11 and 12, the absolute value converting units 13 and 14, the adder 15, the integrating unit 16, and the lens control unit 17 constitute a focus control device.
　撮像部１０は、フォーカスレンズ１０ＦＬ等を含んで構成される光学系と、この光学系を通して撮像素子（図示せず）に結像される被写体を光電変換して映像信号を生成する撮像信号処理部（図示せず）等を備える。 The imaging unit 10 includes an optical system including a focus lens 10FL and the like, and an imaging signal processing unit that generates a video signal by photoelectrically converting a subject imaged on an imaging element (not shown) through the optical system. (Not shown).
　まず、映像ノイズ低減装置について説明する。撮像部１０は、生成した映像信号を、２次元ＨＰＦ１１，１２、ＬＰＦ２１、遅延回路部２２に供給する。映像信号は種々の映像ノイズを含む。 First, the video noise reduction device will be described. The imaging unit 10 supplies the generated video signal to the two- dimensional HPFs 11 and 12, the LPF 21, and the delay circuit unit 22. The video signal includes various video noises.
　ＬＰＦ２１は、入力された映像信号における注目画素と、注目画素の周辺に位置する周辺画素とを含む一定領域の複数の画素の信号を加算平均して、予め設定したカットオフ周波数以下の映像ノイズ成分が抑制された低域信号を出力する。 The LPF 21 averages the signals of a plurality of pixels in a certain area including the target pixel in the input video signal and the peripheral pixels located around the target pixel, and the video noise component is equal to or lower than a preset cutoff frequency. Outputs a low-frequency signal with suppressed.
　遅延回路部２２は、入力された映像信号を、ＬＰＦ２１で発生する遅延時間と同じ時間遅延させて、ＬＰＦ２１から出力される低域信号と時間合わせされた遅延映像信号を出力する。 The delay circuit unit 22 delays the input video signal by the same time as the delay time generated by the LPF 21, and outputs a delayed video signal time-synchronized with the low frequency signal output from the LPF 21.
　２次元ＨＰＦ１１，１２は同じ構成を有するものの、フィルタ特性が異なっている。図２は、２次元ＨＰＦ１１のフィルタ係数の一例を示す。図３は、２次元ＨＰＦ１２のフィルタ係数の一例を示す。 Although the two- dimensional HPFs 11 and 12 have the same configuration, the filter characteristics are different. FIG. 2 shows an example of the filter coefficient of the two-dimensional HPF 11. FIG. 3 shows an example of the filter coefficient of the two-dimensional HPF 12.
　図２，図３に示すように、２次元ＨＰＦ１１，１２には、水平５画素、垂直５画素の２５画素の平面領域における中央の画素を注目画素とし、２５画素それぞれに対する係数が設定されている。 As shown in FIGS. 2 and 3, the two- dimensional HPFs 11 and 12 have coefficients set for each of the 25 pixels, with a central pixel in a 25-pixel planar area of 5 horizontal pixels and 5 vertical pixels as a target pixel. .
　図２に示すように、２次元ＨＰＦ１１は、縦または横方向のテクスチャ・パターンに対して感度が高く、斜め方向のテクスチャ・パターンに対して感度が低いフィルタ特性を有する。２次元ＨＰＦ１１は、縦方向及び横方向の双方のテクスチャ・パターンに対して感度が高いフィルタ特性を有していてもよい。 As shown in FIG. 2, the two-dimensional HPF 11 has a filter characteristic that is highly sensitive to a texture pattern in a vertical or horizontal direction and low in sensitivity to a texture pattern in an oblique direction. The two-dimensional HPF 11 may have a filter characteristic that is highly sensitive to both the vertical and horizontal texture patterns.
　図３に示すように、２次元ＨＰＦ１２は、斜め方向のテクスチャ・パターンに対して感度が高く、縦または横方向のテクスチャ・パターンに対して感度が低いフィルタ特性を有する。 As shown in FIG. 3, the two-dimensional HPF 12 has a filter characteristic that is highly sensitive to a texture pattern in an oblique direction and low in sensitivity to a texture pattern in a vertical or horizontal direction.
　図２及び図３に示すフィルタ係数は、演算量を小さくするために、１，０，－１の３種類の整数のみで構成され、総和は０である。よって、映像信号の注目画素を中心とした２５画素の画像が、輝度値が一様に平坦で映像ノイズのみ重畳されているときには、２次元ＨＰＦ１１，１２によって検出される相関検出値は０またはほぼ０となる。 The filter coefficients shown in FIGS. 2 and 3 are composed of only three kinds of integers of 1, 0, and −1 in order to reduce the amount of calculation, and the sum is zero. Therefore, when a 25-pixel image centered on the target pixel of the video signal has a uniformly flat luminance value and only video noise is superimposed, the correlation detection value detected by the two- dimensional HPFs 11 and 12 is 0 or almost 0.
　図１において、２次元ＨＰＦ１１，１２は、入力された映像信号における注目画素を中心とした２５画素の画像に対し、上記のようなテクスチャ・パターンを検出するための感度の高低の方向が互いに異なるフィルタ特性を有するフィルタを用いたフィルタ演算を施す。２次元ＨＰＦ１１，１２は、フィルタ演算の結果、数値化された相関検出結果を出力する。 In FIG. 1, the two- dimensional HPFs 11 and 12 have different sensitivity levels for detecting a texture pattern as described above with respect to an image of 25 pixels centered on a target pixel in an input video signal. A filter operation using a filter having filter characteristics is performed. The two- dimensional HPFs 11 and 12 output numerical correlation detection results as a result of the filter operation.
　２次元ＨＰＦ１１は、縦または横方向の画像の輝度差がノイズレベルに相当する所定のレベル以下で、縦または横方向のテクスチャ・パターンに対して感度が高いフィルタ特性を有する。よって、２次元ＨＰＦ１１は、入力された２５画素の画像が有するテクスチャ成分が、ノイズレベルと同等の縦または横方向のテクスチャ・パターンと一致するとき、最も大きい値の相関検出結果を出力する。 The two-dimensional HPF 11 has a filter characteristic in which the luminance difference between the vertical and horizontal images is equal to or lower than a predetermined level corresponding to the noise level, and the sensitivity is high with respect to the vertical or horizontal texture pattern. Therefore, the two-dimensional HPF 11 outputs the correlation detection result having the largest value when the texture component included in the input 25-pixel image matches the texture pattern in the vertical or horizontal direction equivalent to the noise level.
　２次元ＨＰＦ１２は、斜め方向の画像の輝度差がノイズレベルに相当する所定のレベル以下で、斜め方向のテクスチャ・パターンに対して感度が高いフィルタ特性を有する。よって、２次元ＨＰＦ１２は、入力された２５画素の画像が有するテクスチャ成分が、ノイズレベルと同等の斜め方向のテクスチャ・パターンと一致するとき、最も大きい値の相関検出結果を出力する。 The two-dimensional HPF 12 has a filter characteristic in which the luminance difference of the image in the oblique direction is not more than a predetermined level corresponding to the noise level and the sensitivity to the texture pattern in the oblique direction is high. Therefore, the two-dimensional HPF 12 outputs the correlation detection result having the largest value when the texture component included in the input 25-pixel image matches the texture pattern in the oblique direction equivalent to the noise level.
　このように、２次元ＨＰＦ１１，１２は、注目画素と前記注目画素の周辺に位置する周辺画素とを含む一定領域の画像に対してフィルタ演算するフィルタ演算部である。２次元ＨＰＦ１１，１２は、ノイズレベルに相当する所定のレベル以下のテクスチャ成分との相関を検出するように設定され、テクスチャ成分を検出するための感度の高低の方向が互いに異なる複数のフィルタ特性を有する複数のフィルタを用いてフィルタ演算する。 As described above, the two- dimensional HPFs 11 and 12 are filter arithmetic units that perform a filter operation on an image in a certain region including the target pixel and peripheral pixels located around the target pixel. The two- dimensional HPFs 11 and 12 are set so as to detect a correlation with a texture component equal to or lower than a predetermined level corresponding to the noise level, and have a plurality of filter characteristics having different sensitivity levels for detecting the texture component. Filter calculation is performed using a plurality of filters.
　２次元ＨＰＦ１１，１２は、フィルタ演算によって、一定領域の画像が有するテクスチャ成分と複数のフィルタ特性それぞれとの複数の相関検出値を求める。 The two- dimensional HPFs 11 and 12 obtain a plurality of correlation detection values between a texture component included in an image in a certain region and each of a plurality of filter characteristics by a filter operation.
　絶対値化部１３は、２次元ＨＰＦ１１から出力された相関検出値を絶対値化し、第１の絶対値信号を加算器１５に供給する。絶対値化部１４は、２次元ＨＰＦ１２から出力された相関検出値を絶対値化し、第２の絶対値信号を加算器１５に供給する。 The absolute value converting unit 13 converts the correlation detection value output from the two-dimensional HPF 11 into an absolute value, and supplies the first absolute value signal to the adder 15. The absolute value converting unit 14 converts the correlation detection value output from the two-dimensional HPF 12 into an absolute value, and supplies the second absolute value signal to the adder 15.
　加算器１５は、第１の絶対値信号と第２の絶対値信号とを加算し、加算信号を混合比計算部２６に供給する。加算器１５は、第１の絶対値信号と第２の絶対値信号とを単純に加算するのではなく、第１の絶対値信号と第２の絶対値信号とのうちの大きい方の絶対値信号を選択する構成としてもよい。 The adder 15 adds the first absolute value signal and the second absolute value signal, and supplies the addition signal to the mixing ratio calculation unit 26. The adder 15 does not simply add the first absolute value signal and the second absolute value signal, but the larger absolute value of the first absolute value signal and the second absolute value signal. It is good also as a structure which selects a signal.
　混合比計算部２６は、加算器１５より出力された加算信号が示す加算値に応じた混合比αを出力する。 The mixing ratio calculation unit 26 outputs a mixing ratio α corresponding to the addition value indicated by the addition signal output from the adder 15.
　図４は、混合比計算部２６の入出力特性の一例であり、横軸は入力信号の値（すなわち加算値）、縦軸は出力信号（すなわち混合比α）の値を示す。図４に示すように、混合比計算部２６は、加算値が０から閾値ＴＨ１１までの範囲では、混合比αとして０を出力する。 FIG. 4 is an example of input / output characteristics of the mixing ratio calculator 26, where the horizontal axis indicates the value of the input signal (ie, the added value) and the vertical axis indicates the value of the output signal (ie, the mixing ratio α). As shown in FIG. 4, the mixture ratio calculation unit 26 outputs 0 as the mixture ratio α in the range from 0 to the threshold value TH11.
　混合比計算部２６は、加算値が閾値ＴＨ１２以上の範囲では、混合比αとして１を出力する。混合比計算部２６は、加算値が閾値ＴＨ１１から閾値ＴＨ１２までの範囲では、混合比αとして０から１までの範囲を直線状に増加する値を出力する。閾値ＴＨ１１，ＴＨ１２は、所望のノイズ低減効果が得られるように実験等に基づいて適切に設定される。 The mixing ratio calculation unit 26 outputs 1 as the mixing ratio α in the range where the addition value is equal to or greater than the threshold value TH12. The mixing ratio calculator 26 outputs a value that linearly increases the range from 0 to 1 as the mixing ratio α in the range where the addition value is from the threshold value TH11 to the threshold value TH12. The thresholds TH11 and TH12 are appropriately set based on experiments or the like so that a desired noise reduction effect can be obtained.
　混合比計算部２６は、図４に示す入出力特性に従って計算した混合比αを乗算器２３，２４に供給する。乗算器２３は、ＬＰＦ２１より出力された低域信号に（１－α）を乗じて出力する。乗算器２４は、遅延回路部２２より出力された遅延映像信号のαを乗じて出力する。加算器２５は、乗算器２３，２４の出力信号を加算して出力する。 The mixing ratio calculation unit 26 supplies the mixing ratio α calculated according to the input / output characteristics shown in FIG. The multiplier 23 multiplies the low frequency signal output from the LPF 21 by (1-α) and outputs the result. The multiplier 24 multiplies α of the delayed video signal output from the delay circuit unit 22 and outputs the result. The adder 25 adds the output signals of the multipliers 23 and 24 and outputs the result.
　混合比αが１であるとき、２５画素の画像はテクスチャ成分が主体の画像であるとみなすことができる。混合比αが１であるとき、乗算器２３の出力信号は０である。よって、映像ノイズ低減装置は、遅延回路部２２より出力された遅延映像信号をそのまま出力する。映像ノイズ低減装置が遅延映像信号をそのまま出力することは、解像度を優先することを意味する。 When the mixing ratio α is 1, an image of 25 pixels can be regarded as an image mainly composed of texture components. When the mixing ratio α is 1, the output signal of the multiplier 23 is 0. Therefore, the video noise reduction device outputs the delayed video signal output from the delay circuit unit 22 as it is. If the video noise reduction device outputs the delayed video signal as it is, it means that the resolution is prioritized.
　混合比αが０であるとき、２５画素の画像は輝度値が平坦な部分に映像ノイズが重畳している画像であるとみなすことができる。混合比αが０であるとき、乗算器２４の出力信号は０である。よって、映像ノイズ低減装置は、ＬＰＦ２１より出力された低域信号をそのまま出力する。ＬＰＦ２１より出力された低域信号をそのまま出力することは、ノイズ低減効果を優先することを意味する。 When the mixing ratio α is 0, an image of 25 pixels can be regarded as an image in which video noise is superimposed on a portion having a flat luminance value. When the mixing ratio α is zero, the output signal of the multiplier 24 is zero. Therefore, the video noise reduction device outputs the low frequency signal output from the LPF 21 as it is. Outputting the low-frequency signal output from the LPF 21 as it is means giving priority to the noise reduction effect.
　映像ノイズ低減装置は、混合比αが０から１までの範囲では、ＬＰＦ２１より出力された低域信号と遅延回路部２２より出力された遅延映像信号とを混合比αの値に応じて混合した混合信号を出力する。両者の信号を混合することにより、映像ノイズとテクスチャ成分とを誤検出しても、画質劣化を目立ちにくくすることができる。 The video noise reduction device mixes the low frequency signal output from the LPF 21 and the delayed video signal output from the delay circuit unit 22 in accordance with the value of the mixing ratio α when the mixing ratio α is in the range from 0 to 1. Output mixed signal. By mixing both signals, even if video noise and texture components are erroneously detected, image quality deterioration can be made inconspicuous.
　映像ノイズ低減装置は、ハードウェア規模をさほど大きくすることなく、映像信号のテクスチャ部分の画質劣化を抑えつつ、画像が平坦な部分に含まれる映像ノイズを低減させることができる。 The video noise reduction device can reduce the video noise contained in the flat portion of the image while suppressing the image quality deterioration of the texture portion of the video signal without increasing the hardware scale.
　映像ノイズ低減装置は、エンコード情報が得られない映像信号であっても、映像ノイズを低減させることができる。映像ノイズ低減装置は、特定方向だけのテクスチャ成分を含む映像信号であっても、擬似ノイズを大幅に抑圧することができる。 The video noise reduction device can reduce video noise even for video signals for which encoding information cannot be obtained. The video noise reduction device can significantly suppress pseudo-noise even for a video signal including a texture component only in a specific direction.
　次に、フォーカス制御装置について説明する。フォーカス制御装置を構成するブロックのうち、映像ノイズ低減装置を構成するブロックと共用されていて、既に説明されたブロックの説明は省略する。 Next, the focus control device will be described. Of the blocks constituting the focus control device, these blocks are shared with the blocks constituting the video noise reduction device, and description of the blocks already described is omitted.
　積算部１６は、加算器１５から画素ごとに出力される加算信号を所定の領域分積算する。所定の領域とは、１フレームの全体、または、１フレーム内の部分的な領域である。積算部１６は、積算値をフォーカス評価値としてレンズ制御部１７に供給する。 The integration unit 16 integrates the addition signal output from the adder 15 for each pixel for a predetermined area. The predetermined area is the whole of one frame or a partial area within one frame. The integrating unit 16 supplies the integrated value to the lens control unit 17 as a focus evaluation value.
　レンズ制御部１７は、入力されたフォーカス評価値に基づいて、撮像部１０が有するフォーカスレンズ１０ＦＬの位置を合焦位置となるように制御する。 The lens control unit 17 controls the position of the focus lens 10FL included in the imaging unit 10 to be the in-focus position based on the input focus evaluation value.
　具体的には、レンズ制御部１７は、フォーカスレンズ１０ＦＬをレンズ可動範囲内で少しずつ移動させ、フォーカス評価値の大きさを監視する。レンズ制御部１７は、フォーカス評価値が最大となるときのフォーカスレンズ１０ＦＬの位置を合焦位置と判断する。レンズ制御部１７は、フォーカスレンズ１０ＦＬの位置が常に合焦位置となるように、フォーカスレンズ１０ＦＬを制御する。 Specifically, the lens control unit 17 moves the focus lens 10FL little by little within the lens movable range, and monitors the magnitude of the focus evaluation value. The lens control unit 17 determines the position of the focus lens 10FL when the focus evaluation value is maximized as the focus position. The lens control unit 17 controls the focus lens 10FL so that the position of the focus lens 10FL is always the in-focus position.
　フォーカス制御値を生成するための２次元ＨＰＦ１１は、前述のように、縦または横方向のテクスチャ・パターンに対して感度が高く、斜め方向のテクスチャ・パターンに対して感度が低い。フォーカス制御値を生成するための２次元ＨＰＦ１２は、斜め方向のテクスチャに対して感度が高く、縦または横方向のテクスチャ・パターンに対して感度が低い。 As described above, the two-dimensional HPF 11 for generating the focus control value has high sensitivity to the vertical or horizontal texture pattern and low sensitivity to the diagonal texture pattern. The two-dimensional HPF 12 for generating the focus control value has high sensitivity to the texture in the oblique direction and low sensitivity to the texture pattern in the vertical or horizontal direction.
　２次元ＨＰＦ１１，１２は、いずれも、ランダムノイズに対して感度が極めて低い。２次元ＨＰＦ１１，１２のフィルタ係数は演算量が小さくなるように－１，０，１のみで構成され、総和は０である。 The two- dimensional HPFs 11 and 12 are extremely low in sensitivity to random noise. The filter coefficients of the two- dimensional HPFs 11 and 12 are composed of only −1, 0, and 1 so that the calculation amount is small, and the sum is zero.
　従って、本実施形態のフォーカス制御装置は、比較的簡単な構成で、合焦の検出感度が高い。本実施形態のフォーカス制御装置によれば、ノイズの影響を受けにくいオートフォーカスが可能となる。 Therefore, the focus control device of this embodiment has a relatively simple configuration and high focus detection sensitivity. According to the focus control apparatus of the present embodiment, autofocus that is less susceptible to noise is possible.
　本実施形態のフォーカス制御装置は、２次元ＨＰＦ１１、２次元ＨＰＦ１２、絶対値化部１３、絶対値化部１４、加算器１５を、映像ノイズ低減装置と共用する構成としている。よって、フォーカス制御装置の大幅なコストダウンが可能である。 The focus control device of this embodiment is configured to share the two-dimensional HPF 11, the two-dimensional HPF 12, the absolute value converting unit 13, the absolute value converting unit 14, and the adder 15 with the video noise reducing device. Therefore, the cost of the focus control device can be greatly reduced.
　しかしながら、２次元ＨＰＦ１１、２次元ＨＰＦ１２、絶対値化部１３、絶対値化部１４、加算器１５を映像ノイズ低減装置と共用せず、フォーカス制御装置専用の構成として設けてもよい。このような構成であっても、比較的簡単な構成で、合焦の検出感度が高く、かつ、ノイズの影響を受けにくいオートフォーカスが可能となるという効果を奏する。 However, the two-dimensional HPF 11, the two-dimensional HPF 12, the absolute value conversion unit 13, the absolute value conversion unit 14, and the adder 15 may not be shared with the video noise reduction device but may be provided as a configuration dedicated to the focus control device. Even with such a configuration, there is an effect that auto-focusing is possible with a relatively simple configuration, high focus detection sensitivity, and hardly affected by noise.
＜第２実施形態＞
　図５に示す撮像装置２００において、図１に示す撮像装置１００と同一部分には同一符号を付し、同一部分の説明を省略することがある。撮像装置２００における映像ノイズ低減装置の構成及び動作は、撮像装置１００におけるそれと同じである。 Second Embodiment
In the imaging apparatus 200 shown in FIG. 5, the same parts as those of the imaging apparatus 100 shown in FIG. The configuration and operation of the video noise reduction device in the imaging device 200 are the same as those in the imaging device 100.
　図５において、撮像装置２００は、撮像装置１００が備える構成の他に、周波数領域フィルタ３０、絶対値化部３１、混合比計算部３２、乗算器３３，３４、加算器３５を備える。 5, the imaging apparatus 200 includes a frequency domain filter 30, an absolute value conversion unit 31, a mixture ratio calculation unit 32, multipliers 33 and 34, and an adder 35 in addition to the configuration included in the imaging apparatus 100.
　図５において、撮像部１０は、生成した映像信号を、周波数領域フィルタ３０に供給する。周波数領域フィルタ３０は、所定の周波数領域を抽出するバンドパスフィルタ（ＢＰＦ）またはＨＰＦよりなる。図６は、周波数領域フィルタ３０が有するフィルタ特性の一例である。 In FIG. 5, the imaging unit 10 supplies the generated video signal to the frequency domain filter 30. The frequency domain filter 30 is composed of a band pass filter (BPF) or HPF that extracts a predetermined frequency domain. FIG. 6 is an example of the filter characteristics that the frequency domain filter 30 has.
　絶対値化部３１は、周波数領域フィルタ３０の出力を絶対化する。周波数領域フィルタ３０の出力は、低コントラストのときには小さな値となり、高コントラストのときには大きな値となる。絶対値化部３１より出力された絶対値信号は、混合比計算部３２及び乗算器３４に供給される。 The absolute value conversion unit 31 absoluteizes the output of the frequency domain filter 30. The output of the frequency domain filter 30 has a small value when the contrast is low, and a large value when the contrast is high. The absolute value signal output from the absolute value converting unit 31 is supplied to the mixture ratio calculating unit 32 and the multiplier 34.
　混合比計算部３２は、入力信号（絶対値化部３１からの絶対値信号）と閾値ＴＨ２１，ＴＨ２２とを比較する。混合比計算部３２は、図７に示すように、絶対値信号が０から閾値ＴＨ２１までの範囲では、混合比βとして０を出力する。混合比計算部２６は、絶対値信号が閾値ＴＨ２１以上の範囲では、混合比βとして１を出力する。 The mixing ratio calculation unit 32 compares the input signal (the absolute value signal from the absolute value conversion unit 31) with the threshold values TH21 and TH22. As shown in FIG. 7, the mixture ratio calculation unit 32 outputs 0 as the mixture ratio β in the range where the absolute value signal is from 0 to the threshold value TH21. The mixing ratio calculator 26 outputs 1 as the mixing ratio β in the range where the absolute value signal is equal to or greater than the threshold value TH21.
　混合比計算部３２は、絶対値信号が閾値ＴＨ２１から閾値ＴＨ２２までの範囲では、混合比βとして０から１までの範囲を直線状に増加する値を出力する。 The mixing ratio calculation unit 32 outputs a value that linearly increases the range from 0 to 1 as the mixing ratio β when the absolute value signal is in the range from the threshold value TH21 to the threshold value TH22.
　混合比計算部３２は、図７に示す入出力特性に従って計算した混合比βを乗算器３３，３４に供給する。乗算器３３は、加算器１５より出力された加算信号に（１－β）を乗じて出力する。乗算器３４は、絶対値化部３１より出力された絶対値信号にβを乗じて出力する。加算器３５は、乗算器３３，３４の出力信号を加算して出力する。 The mixing ratio calculation unit 32 supplies the mixing ratio β calculated according to the input / output characteristics shown in FIG. The multiplier 33 multiplies the added signal output from the adder 15 by (1-β) and outputs the result. The multiplier 34 multiplies the absolute value signal output from the absolute value converting unit 31 by β and outputs the result. The adder 35 adds the output signals of the multipliers 33 and 34 and outputs the result.
　第１実施形態と同様に、積算部１６は、加算器３５から画素ごとに出力される加算信号を所定の領域分積算し、積算値をフォーカス評価値としてレンズ制御部１７に供給する。レンズ制御部１７は、入力されたフォーカス評価値に基づいて、撮像部１０が有するフォーカスレンズ１０ＦＬの位置を合焦位置となるように制御する。 As in the first embodiment, the integration unit 16 integrates the addition signal output for each pixel from the adder 35 for a predetermined area, and supplies the integration value to the lens control unit 17 as a focus evaluation value. The lens control unit 17 controls the position of the focus lens 10FL included in the imaging unit 10 to be the in-focus position based on the input focus evaluation value.
　所定の周波数領域を抽出する周波数領域フィルタ３０は、低コントラストかつ画像が周波数領域フィルタ３０の通過領域の周波数成分を含む場合、映像ノイズとテクスチャ成分との分離が困難な場合がある。 When the frequency domain filter 30 that extracts a predetermined frequency domain has low contrast and the image includes the frequency component of the pass domain of the frequency domain filter 30, it may be difficult to separate the video noise from the texture component.
　そこで、第２実施形態においては、低コントラスト時には相関検出フィルタである２次元ＨＰＦ１１，１２の出力を使用し、高コントラスト時には周波数領域フィルタ３０の出力を使用するように切り替えるか、両者の混合比を変更するように構成している。これによって、低コントラスト時に、映像ノイズが周波数領域フィルタ３０の出力に影響を与えるのを抑制している。 Therefore, in the second embodiment, the outputs of the two- dimensional HPFs 11 and 12 that are correlation detection filters are used when the contrast is low, and the output of the frequency domain filter 30 is used when the contrast is high, or the mixing ratio between the two is changed. It is configured to change. This suppresses the influence of the video noise on the output of the frequency domain filter 30 when the contrast is low.
　第２実施形態によれば、周波数領域フィルタ３０の特性を活かしつつ、低コントラスト時の映像ノイズの影響を抑えることができ、フォーカス制御の性能を向上させることができる。 According to the second embodiment, while taking advantage of the characteristics of the frequency domain filter 30, it is possible to suppress the influence of video noise at the time of low contrast and improve the performance of focus control.
　本発明は以上説明した第１，第２実施形態に限定されることはなく、本発明の要旨を逸脱しない範囲において種々変更可能である。図１，図５において、フィルタ演算部を、感度の高低の方向が互いに異なる３つ以上の２次元ＨＰＦを用いて構成してもよい。 The present invention is not limited to the first and second embodiments described above, and various modifications can be made without departing from the scope of the present invention. 1 and 5, the filter calculation unit may be configured using three or more two-dimensional HPFs having different sensitivity levels.
　本発明は、図１，図５におけるフォーカス制御装置の各構成の機能をコンピュータにより実現させるコンピュータプログラムを包含する。コンピュータプログラムは、記録媒体からコンピュータに取り込まれてもよいし、ネットワークを介して配信されてコンピュータにダウンロードされてもよい。 The present invention includes a computer program for realizing the functions of the respective components of the focus control device in FIGS. 1 and 5 by a computer. The computer program may be taken into the computer from a recording medium, or distributed via a network and downloaded to the computer.
　本発明は、コントラスト方式のフォーカス制御装置に利用できる。 The present invention can be used for a contrast type focus control device.

Claims (4)

  1. 　フォーカスレンズを有する撮像部と、
    　前記撮像部から出力された映像信号における注目画素と前記注目画素の周辺に位置する周辺画素とを含む一定領域の画像に対し、ノイズレベルに相当する所定のレベル以下のテクスチャ成分との相関を検出するように設定され、テクスチャ成分を検出するための感度の高低の方向が互いに異なる複数のフィルタ特性を有する複数のフィルタを用いてフィルタ演算して、前記一定領域の画像が有するテクスチャ成分と前記複数のフィルタ特性それぞれとの複数の相関検出値を求めるフィルタ演算部と、
    　前記複数の相関検出値それぞれを絶対値化する絶対値化部と、
    　前記絶対値化部によって絶対値化された前記複数の相関検出値を加算する加算器と、
    　前記加算器より出力された加算値を所定の領域内で積算する積算部と、
    　前記積算部より出力された積算値をフォーカス評価値とし、前記フォーカス評価値に応じて前記フォーカスレンズの位置を制御するレンズ制御部と、
    　を備えることを特徴とするフォーカス制御装置。     An imaging unit having a focus lens;
    Detects correlation between a target area in a video signal output from the image pickup unit and a peripheral region located around the target pixel, and a texture component below a predetermined level corresponding to a noise level. A plurality of filters having a plurality of filter characteristics whose sensitivity levels for detecting texture components are different from each other. A filter calculation unit for obtaining a plurality of correlation detection values with each of the filter characteristics of
    An absolute value converting unit that converts each of the plurality of correlation detection values into an absolute value;
    An adder for adding the plurality of correlation detection values converted into absolute values by the absolute value conversion unit;
    An accumulator for accumulating the added value output from the adder within a predetermined area;
    The integrated value output from the integrating unit is a focus evaluation value, and a lens control unit that controls the position of the focus lens according to the focus evaluation value;
    A focus control device comprising:
  2. 　前記フィルタ演算部は、
    　斜め方向より縦または横方向に感度が高い第１の係数パターンを有する第１のフィルタ係数を用いてフィルタ演算する第１の２次元ハイパスフィルタと、
    　縦または横方向より斜め方向に感度が高い第２の係数パターンを有する第２のフィルタ係数を用いてフィルタ演算する第２の２次元ハイパスフィルタと、
    　を有する
    　ことを特徴とする請求項１記載のフォーカス制御装置。     The filter operation unit
    A first two-dimensional high-pass filter that performs a filter operation using a first filter coefficient having a first coefficient pattern having higher sensitivity in the vertical or horizontal direction than in the oblique direction;
    A second two-dimensional high-pass filter that performs a filter operation using a second filter coefficient having a second coefficient pattern that is more sensitive in a diagonal direction than the vertical or horizontal direction;
    The focus control device according to claim 1, comprising:
  3. 　前記第１及び第２のフィルタ係数はそれぞれ係数の総和が０であることを特徴とする請求項２記載のフォーカス制御装置。 3. The focus control apparatus according to claim 2, wherein the first and second filter coefficients each have a sum of coefficients of 0.
  4. 　撮像部から出力された映像信号における注目画素と前記注目画素の周辺に位置する周辺画素とを含む一定領域の画像に対し、ノイズレベルに相当する所定のレベル以下のテクスチャ成分との相関を検出するように設定され、テクスチャ成分を検出するための感度の高低の方向が互いに異なる複数のフィルタ特性を有する複数のフィルタを用いてフィルタ演算して、前記一定領域の画像が有するテクスチャ成分と前記複数のフィルタ特性それぞれとの複数の相関検出値を求め、
    　前記複数の相関検出値それぞれを絶対値化して複数の絶対値信号を生成し、
    　前記複数の絶対値信号を加算して加算信号を生成し、
    　前記加算信号を所定の領域内で積算して積算値を生成し、
    　前記積算値をフォーカス評価値とし、前記フォーカス評価値に応じて前記撮像部が備えるフォーカスレンズの位置を制御する
    　ことを特徴とするフォーカス制御方法。     A correlation between a target region in the video signal output from the imaging unit and a certain region including a peripheral pixel located around the target pixel is detected with a texture component having a level equal to or lower than a predetermined level corresponding to a noise level. And performing a filter operation using a plurality of filters having a plurality of filter characteristics having different sensitivity levels for detecting the texture component, and the texture component included in the image of the certain region and the plurality of the plurality of filters. Find multiple correlation detection values with each filter characteristic,
    Each of the plurality of correlation detection values is converted into an absolute value to generate a plurality of absolute value signals,
    Adding the plurality of absolute value signals to generate an addition signal;
    The added signal is integrated within a predetermined area to generate an integrated value,
    A focus control method comprising: setting the integrated value as a focus evaluation value, and controlling a position of a focus lens included in the imaging unit according to the focus evaluation value.
PCT/JP2014/068200 2013-07-16 2014-07-08 Focus control device, and focus control method WO2015008663A1 (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
JP2013147453 2013-07-16
JP2013-147453 2013-07-16
JP2014-072386 2014-03-31
JP2014072386A JP2015038593A (en) 2013-07-16 2014-03-31 Focus control device and focus control method

Publications (1)

Publication Number Publication Date
WO2015008663A1 true WO2015008663A1 (en) 2015-01-22

Family

ID=52346130

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2014/068200 WO2015008663A1 (en) 2013-07-16 2014-07-08 Focus control device, and focus control method

Country Status (2)

Country Link
JP (1) JP2015038593A (en)
WO (1) WO2015008663A1 (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07170395A (en) * 1993-10-08 1995-07-04 Matsushita Electric Ind Co Ltd Area discriminating device and gray level transformation processor
JP2011257769A (en) * 2011-08-02 2011-12-22 Ricoh Co Ltd Imaging apparatus and imaging method using the same
JP2013131862A (en) * 2011-12-20 2013-07-04 Olympus Corp Image processing system and microscope system including the same

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07170395A (en) * 1993-10-08 1995-07-04 Matsushita Electric Ind Co Ltd Area discriminating device and gray level transformation processor
JP2011257769A (en) * 2011-08-02 2011-12-22 Ricoh Co Ltd Imaging apparatus and imaging method using the same
JP2013131862A (en) * 2011-12-20 2013-07-04 Olympus Corp Image processing system and microscope system including the same

Also Published As

Publication number Publication date
JP2015038593A (en) 2015-02-26

Similar Documents

Publication Publication Date Title
KR102182695B1 (en) Method and Apparatus for Noise Reduction
JP5948073B2 (en) Image signal processing apparatus and image signal processing method
JP6056702B2 (en) Image processing apparatus, image processing method, and program
KR20140070216A (en) Method and Apparatus for processing image
WO2013031102A1 (en) Image processing device, image capture device, and image processing method
JP2007041834A (en) Image processor
EP3745704B1 (en) Image processing device, output information control method, and program
JP2015034869A5 (en)
US9807325B2 (en) Imaging apparatus, control method for imaging apparatus, and non-transitory computer-readable storage medium
JP2014021928A (en) Image processor, image processing method and program
JP2010268426A (en) Image processing apparatus, image processing method, and program
JP5291788B2 (en) Imaging device
JPWO2017154293A1 (en) Image processing apparatus, imaging apparatus, image processing method, and program
JP2009194721A (en) Image signal processing device, image signal processing method, and imaging device
US9007494B2 (en) Image processing apparatus, method for controlling the same and storage medium
JP2015177528A (en) image processing apparatus, image processing method, and program
WO2015008663A1 (en) Focus control device, and focus control method
KR101781766B1 (en) Video signal noise elimination circuit and video signal noise elimination method
US20180352153A1 (en) Image processing apparatus that generates evaluation value for evaluating luminance of image data, image capturing apparatus, method for processing image, and program
WO2013150824A1 (en) Image processing device and method, and image processing program
JP6029464B2 (en) Imaging device, control method thereof, and control program
JP2019140605A (en) Image processing system, image processing method and program
JP5242750B2 (en) Image processing device
JP5506623B2 (en) Video processing apparatus and video processing method
US11019348B2 (en) Image processing apparatus and image processing method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14826040

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14826040

Country of ref document: EP

Kind code of ref document: A1