JP2011059147A - Method and device for evaluating image blur - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To evaluate whether or not an image blur has occurred based on a fixed criteria regardless of the size of an image to be evaluated. <P>SOLUTION: The method for evaluating the image blur evaluates the image blur caused on an image to be evaluated, which is photographed by an image pickup device. The method includes: a take-in step 102 of taking in an image to be evaluated; a frame setting step 103 of setting an evaluation frame to a region of interest in the image to be evaluated for the image taken in the take-in process; a position setting step 104 of setting an evaluation image position in the evaluation frame; an operation step 105 of performing evaluation image operation for a partial image in the evaluation image position set in the position setting step; a determination step 106 of determining the image blur based on the image blur evaluation value calculated in the operation step; and a result output step of outputting the determination result of the determination step. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to an image shake evaluation method and an image shake evaluation apparatus for evaluating shake of a photographed image.

  In an imaging apparatus, an image of a subject is often shaken by camera shake or vibration, and the image is often difficult to see. Recently, a high-magnification lens has been adopted, and image blurring is particularly noticeable especially on the telephoto side. Many image pickup apparatuses having an apparatus that corrects image shake for such camera shake have been proposed. In this type of imaging apparatus, the number of steps of shutter speed is generally used as an index indicating the effect of image blur correction. In order to calculate the number of shutter speed steps indicating the effect of image blur correction, image blur is evaluated. It is requested to do.

  Many methods and apparatuses for evaluating the image quality of an image formed on an image quality forming medium such as paper have been proposed. Image quality evaluation methods include image quality psychological evaluation and physical evaluation. The image quality psychological evaluation is a so-called sensory evaluation in which the degree of “beauty” and “preferability” that a person feels psychologically by visually observing an image is quantified. In the physical evaluation, attention is paid to each of a plurality of image quality factors (image quality evaluation items) that influence the image quality, for example, gradation and color reproduction range, and the evaluation is performed based on objectively measured values of the image structure. Is.

  In the case of performing image quality evaluation in which the image quality psychological evaluation and physical evaluation are combined, an image quality evaluation method and apparatus have been proposed in which the correlation between both evaluation results is good (for example, Patent Document 1).

JP 10-63859 A

  However, in the above-mentioned patent document 1, image quality evaluation that combines image quality psychological evaluation and image quality physical evaluation is performed, and image blur cannot be evaluated.

(Object of invention)
SUMMARY OF THE INVENTION An object of the present invention is to provide an image shake evaluation method and an image shake evaluation apparatus capable of evaluating whether an image has no shake or an image with shake on a constant basis regardless of the size of the image to be evaluated. It is something to be offered.

  In order to achieve the above object, an image shake evaluation method of the present invention is an image shake evaluation method for evaluating shake of an image to be evaluated photographed by an imaging apparatus, and a capturing step for capturing the image to be evaluated, For the image captured in the capturing step, a frame setting step for setting an evaluation frame for the attention area of the image to be evaluated, a position setting step for setting an evaluation image position in the evaluation frame, and A calculation step of performing an evaluation image calculation on the partial image at the evaluation image position set in the position setting step, a determination step of determining a shake image based on a shake evaluation value calculated in the calculation step; And a result output step of outputting a determination result of the determination step.

  As described above, according to the present invention, it is possible to evaluate whether an image has no shake or an image with shake on a constant basis regardless of the size of the image to be evaluated.

1 is a block diagram illustrating an image blur evaluation apparatus according to Embodiment 1 of the present invention. FIG. 4 is a diagram illustrating an example of a preferable part of an image to be evaluated according to Example 1. 6 is a flowchart illustrating an operation of an evaluation frame setting unit according to the first embodiment. It is a figure which shows the example of the evaluation frame setting in Example 1. FIG. FIG. 6 is a diagram illustrating an example of evaluation image position setting in the first embodiment. 6 is a flowchart illustrating an operation of an evaluation image calculation unit according to the first embodiment. It is a figure which shows the example which normalizes the luminance value of black and white in Example 1. FIG. 6 is a diagram illustrating a luminance value distribution at an evaluation image position in Example 1. FIG. 10 is a flowchart illustrating an operation of an evaluation image calculation unit according to the second embodiment. It is a figure which shows the luminance value distribution in the evaluation image position in Example 2. FIG.

  The mode for carrying out the present invention is as shown in Examples 1 and 2 below.

  FIG. 1 is a block diagram illustrating an overall configuration of an image blur evaluation apparatus according to Embodiment 1 of the present invention. The image blur evaluation apparatus includes a host computer 101 that is a central computer that controls each unit of the image blur evaluation apparatus, an evaluation image capturing unit 102, an evaluation frame setting unit 103, and an evaluation image position setting unit 104. Furthermore, an evaluation image calculation unit 105, an evaluation determination unit 106, and an evaluation result output unit 107 are provided.

  The evaluation image capturing unit 102 captures an evaluation image captured by the imaging device into the image shake evaluation device. Since the image to be evaluated can easily determine the direction of image blur, an image having vertical and horizontal lines is assumed. For example, an image obtained by photographing the charts shown in FIGS. In the image shake evaluation, since it is assumed that an image having a large change in luminance value (chart as shown in FIG. 2) is taken, a general image is not an evaluation target.

  The evaluation frame setting unit 103 detects a partial region (attention region) of the target image to be evaluated among the evaluation images captured by the evaluation image capturing unit 102, and sets an evaluation frame for the detected region. To do. Hereinafter, the operation of setting the evaluation frame in the evaluation frame setting unit 103 will be described using the flowchart of FIG.

  First, in step S301, scanning of the horizontal line of the image to be evaluated is started, and in the next step S303, it is determined whether there is a change in luminance value. If there is no change in luminance value on the horizontal line, the process proceeds from step S302 to step S304 immediately. If there is a luminance value change in the horizontal line, the process proceeds from step S302 to step S303, and the horizontal position coordinate where the luminance value change has occurred is recorded. Then, in the next step S304, it is checked whether scanning has been completed for all horizontal lines of the image to be evaluated. If not, the process returns to step S301, and the same operation is repeated.

  Thereafter, when scanning is completed for all horizontal lines of the image to be evaluated, the process proceeds from step S304 to step S305. This time, scanning of the vertical line of the image to be evaluated is started. In the next step S306, whether there is a change in luminance value or not. Determine. If there is no luminance value change in the vertical line, the process proceeds from step S306 to step S304. If there is a change in the brightness value of the vertical line, the process immediately proceeds from step S306 to step S307, and the vertical position coordinates where the brightness value change has occurred are recorded. Then, in the next step S308, it is checked whether or not scanning has been completed for all the vertical lines of the image to be evaluated. If not, the process returns to step S305 and the same operation is repeated.

  When scanning is completed for all vertical lines of the image to be evaluated, the process advances from step S308 to step S309, and an evaluation frame is set based on the position coordinates of the luminance value changes of all horizontal lines and all vertical lines of the image to be evaluated.

  By setting the evaluation frame through the above processing, the evaluation frame is also shifted from the center of the optical axis of the image with respect to the image to be evaluated as shown in FIGS. 4A and 4B. An appropriate evaluation frame 108 can be set. Note that the horizontal and vertical scanning starts from the upper left position of the image.

  Returning to FIG. 1, the evaluation image position setting unit 104 sets a target evaluation image position within the evaluation frame set by the evaluation frame setting unit 103. As shown in FIG. 5, the evaluation image position is set by paying attention to several points perpendicular to the vertical and horizontal lines. In FIG. 5, for example, 10 evaluation image positions are set on the vertical line and 10 evaluation image positions are set on the horizontal line. However, the evaluation image position is not limited to this position, and can be changed as necessary.

  The evaluation image calculation unit 105 performs evaluation image calculation on the evaluation image position set by the evaluation image position setting unit 104. Hereinafter, the evaluation image calculation in the evaluation image calculation unit 105 will be described with reference to the flowchart of FIG.

  First, in step S601, white and black luminance values are normalized with respect to the image to be evaluated. FIG. 7 shows the positions for normalizing the black and white luminance values in step S601.

  As shown in FIG. 7, a point 201 in the evaluation frame 108 is used as a white reference, and a point 202 outside the evaluation frame is used as a black reference. In FIG. 7, for example, four points within the evaluation frame and four points outside the evaluation frame are used as a reference. However, the number of sample points for normalization is not limited to this number, and can be changed as necessary. In this way, by normalizing the black and white luminance values and setting the luminance values serving as the black and white standards in the image to be evaluated, it is possible to suppress evaluation errors due to the color of the image to be evaluated. In normalization, the luminance average value of four points in the evaluation frame may be used as a white reference, and the luminance average value of four points outside the evaluation frame may be used as a black reference.

  Further, the luminance value distribution at the evaluation image position is as shown in FIG.

Returning to FIG. 6, in step S602, the luminance value distribution as shown in FIG. 8 is divided into a white region above the predetermined level, a black region below the predetermined level, and a gray region below the other portions. Then, in the next step S603, the following formula is applied to the luminance area divided as described above: shake evaluation value = gray area / (gray area + black area)
Is applied to calculate an image shake evaluation value. This shake evaluation value becomes smaller as the gray area becomes smaller. That is, the smaller the shake evaluation value is, the smaller the gray area is, and the boundary between the white area and the black area is emphasized, so that an image with less shake can be obtained.

  In the next step S604, it is determined whether or not a shake evaluation value has been calculated for all evaluation image positions set by the evaluation image position setting unit 104. If no shake evaluation value has been calculated for all evaluation image positions, the process proceeds to step S605, moves to the next evaluation image position, calculates a shake evaluation value, returns to step S602, and performs the same operation. repeat. Thus, the shake evaluation value is calculated for each set evaluation image position.

  If it is determined in step S604 that the shake evaluation values have been calculated for all the evaluation image positions, the process proceeds to step S606, where the calculated shake evaluation values of all the evaluation image positions are summed or averaged, and the image to be evaluated The overall runout evaluation value is calculated. When calculating the shake evaluation value for the entire image to be evaluated, considering the variation of the calculated shake evaluation value, omitting the maximum value or several points close to the maximum value and some points close to the minimum value or minimum value. The shake evaluation value of the entire image to be evaluated may be calculated using several points.

  Returning to FIG. 1, the evaluation determination unit 106 determines whether there is an image shake based on the shake evaluation value calculated by the evaluation image calculation unit 105. That is, an image with an image shake evaluation value smaller than a predetermined value is an image without shake, and an image with an image shake evaluation value larger than a predetermined value is determined as an image with shake. The evaluation result output unit 107 outputs the determination result from the evaluation determination unit 106. Note that the output of the evaluation determination unit 106 is output by, for example, a display or a printer.

  The host computer 101 controls the operation of each unit and stores image data and image processing calculation data in each unit.

  According to the first embodiment described above, the luminance value at the evaluation image position within the evaluation frame of the image to be evaluated is calculated, divided into a white area, a black area, and a gray area, and the ratio of the black area to the gray area is calculated. Image blur evaluation values are used. Therefore, regardless of the size of the image to be evaluated, it is possible to always perform reliable image shake determination on a constant basis, that is, to evaluate whether the image has no shake or an image with shake.

  Next, an image shake evaluation apparatus according to Embodiment 2 of the present invention will be described. The circuit components of the image blur evaluation device are the same as those shown in FIG.

  The evaluation image calculation in the evaluation image calculation unit 105 according to the second embodiment of the present invention will be described using the flowchart of FIG.

  First, in step S901, white and black luminance values are normalized for the image to be evaluated. In this way, by normalizing the black and white luminance values and setting the luminance values serving as the black and white standards in the image to be evaluated, it is possible to suppress evaluation errors due to the color of the image to be evaluated. In the next step S902, with respect to the luminance value distribution at the evaluation image position, a predetermined level or higher is divided into a white area, a predetermined level or lower is divided into a black area, and the other parts are divided into gray areas. In subsequent step S603, the gray region is further divided into several types of gray regions according to the luminance value. The luminance value distribution at the evaluation image position is as shown in FIG. In FIG. 10, for example, the gray area is divided into three areas, but the number of divisions can be selected.

なる式により振れ評価値として算出している。なお、灰色領域をｎ個に分割し、各灰色領域に対する重み付け量をＫｘとする。この振れ評価値は灰色領域が小さいほど小さくなる。つまり、振れ評価値が小さいほど灰色領域が小さく、白色領域と黒色領域の境界が強調されており、振れの少ない画像ということができる。 In the next step S904, a shake evaluation value is calculated by applying the following formula to the luminance region divided as described above. This is because the gray area is divided into a plurality of areas according to the luminance value, and each area is weighted.

As a runout evaluation value, the following formula is used. Note that the gray area is divided into n, and the weighting amount for each gray area is Kx. This shake evaluation value becomes smaller as the gray area becomes smaller. That is, the smaller the shake evaluation value is, the smaller the gray area is, and the boundary between the white area and the black area is emphasized, so that an image with less shake can be obtained.

  In the next step S905, it is determined whether or not a shake evaluation value has been calculated for all evaluation image positions set by the evaluation image position setting unit 104. If no shake evaluation value has been calculated for all the evaluation image positions, the process proceeds to step S906, moves to the next evaluation image position, calculates a shake evaluation value, returns to step S902, and performs the same operation. repeat. Thus, the shake evaluation value is calculated for each set evaluation image position.

  If it is determined in step S905 that the shake evaluation values have been calculated for all the evaluation image positions, the process proceeds to step S907, where the calculated shake evaluation values of all the evaluation image positions are summed or averaged, and evaluated. The shake evaluation value calculation for the entire image is performed. When calculating the shake evaluation value for the entire image to be evaluated, considering the variation of the calculated shake evaluation value, omitting the maximum value or several points close to the maximum value and some points close to the minimum value or minimum value. The shake evaluation value of the entire image to be evaluated may be calculated using several points.

  According to the second embodiment described above, the luminance value at a predetermined position of the image to be evaluated is calculated and divided into a white area, a black area, and a gray area. Further, the gray area is divided into a plurality of areas according to the luminance value, and the ratio of the black area to the gray area is calculated as an image shake evaluation value with respect to a value obtained by weighting the black area and the divided gray area. .

  By performing weighting on the gray area as described above and setting so as to emphasize a part that is felt to be shaken at the time of visual observation, it is possible to perform shake evaluation that is close to a human visual sense.

  According to each of the embodiments described above, the image shake evaluation apparatus or the image shake evaluation method for evaluating the image shake of the image to be evaluated photographed by the image pickup apparatus is used with the following means (or by a process). An evaluation is performed. More specifically, a capturing unit (process) that captures an image to be evaluated, and a frame setting unit (process) that sets an evaluation frame for a region of interest of the evaluated image with respect to the image captured by the capturing unit (process). Have Further, within the evaluation frame, position setting means (step) for setting the evaluation image position, and calculation means (step) for performing evaluation image calculation on the partial image at the evaluation image position set in the position setting means (step). Have Furthermore, it has a determination means (step) for determining a shake image based on the shake evaluation value calculated in the calculation means (step), and a result output means (step) for outputting the determination result of the determination means (step). .

  In the frame setting means (step), an evaluation frame is set according to the position where the horizontal and vertical luminance values of the image to be evaluated change. Further, the calculation means (process) normalizes the luminance value at the evaluation image position using the luminance values inside and outside the evaluation frame. Further, the calculation means (process) divides the image into a plurality of areas according to the luminance value. Further, the calculation means (step) calculates a shake evaluation value obtained by weighting each of the divided areas.

  As described above, the luminance value of the evaluation image position within the evaluation frame of the image to be evaluated is calculated, divided into the white region, the black region, and the gray region, the ratio of the black region and the gray region is calculated, and the shake evaluation value and By doing so, a highly reliable shake determination can always be performed on a constant basis regardless of the size of the image to be evaluated. Specifically, if shake evaluation is performed with the number of pixels, the shake cannot be determined properly due to a change in the angle of view (chart line thickness) that occurs during chart shooting (here, shooting at a certain distance between charts) However, if the shooting position is slightly moved back and forth unintentionally). On the other hand, in this embodiment, since the shake of the image is evaluated by the ratio of the black area and the gray area, even if there is a slight change in the angle of view, the shake determination is not affected.

DESCRIPTION OF SYMBOLS 101 Host computer 102 Evaluation image taking-in part 103 Evaluation frame setting part 104 Evaluation image position setting part 105 Evaluation image calculating part 106 Evaluation determination part 107 Evaluation result output part

Claims (5)

An image shake evaluation method for evaluating shake of an image to be evaluated photographed by an imaging device,
A capturing step for capturing the image to be evaluated;
A frame setting step for setting an evaluation frame for the attention area of the image to be evaluated for the image captured in the capturing step;
Within the evaluation frame, a position setting step for setting an evaluation image position;
A calculation step of performing an evaluation image calculation on the partial image at the evaluation image position set in the position setting step;
A determination step of determining a shake image based on the shake evaluation value calculated in the calculation step;
And a result output step of outputting a determination result of the determination step.   The image blur evaluation method according to claim 1, wherein in the frame setting step, the evaluation frame is set according to a position where a luminance value in a horizontal direction and a vertical direction of the image to be evaluated changes.   In the calculation step, the luminance values at the evaluation image position are normalized using the luminance values inside and outside the evaluation frame, and the distribution of the normalized luminance values is divided into a plurality of luminance levels according to a plurality of levels. The image blur evaluation method according to claim 1, wherein the blur evaluation value is calculated based on the divided luminance area.   5. The image blur evaluation method according to claim 4, wherein, in the calculation step, the blur evaluation value is calculated by weighting each of the plurality of divided luminance regions. An image shake evaluation apparatus for evaluating shake of an image to be evaluated photographed by an imaging apparatus,
Capture means for capturing the image to be evaluated;
Frame setting means for setting an evaluation frame for the attention area of the image to be evaluated for the image captured by the capturing means;
Position setting means for setting an evaluation image position within the evaluation frame;
A calculation means for performing an evaluation image calculation on a partial image at the evaluation image position set by the position setting means;
Determination means for determining a shake image based on a shake evaluation value calculated by the calculation means;
An image blur evaluation apparatus comprising: a result output unit that outputs a determination result of the determination unit.

Images