JP2009188624A - Image processor and electronic camera, and allowable value setting method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the following problem in a conventional image processing technology for correcting the pixel value of defective pixel: an appropriate corrected pixel value is not always obtained and the image quality of photographed images is damaged at a part where an image structure is flat and a part provided with high frequency components. <P>SOLUTION: The image processor includes: a first corrected pixel value calculation part for calculating a prescribed correction amount for the pixel value of the defective pixel of an imaging device and calculating a first corrected pixel value; a corrected pixel value allowable range setting part for setting a corrected pixel value allowable range including the first corrected pixel value; a second corrected pixel value calculation part for performing an interpolating operation using the pixel values of pixels around the defective pixel and calculating a second corrected pixel value; a third corrected pixel value calculation part for comparing the second corrected pixel value with the corrected pixel value allowable range and calculating a third corrected pixel value; and a correction value substituting part for replacing the pixel value of the defective pixel by the third corrected pixel value as. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing technique for correcting pixel values of defective pixels of an image sensor.

  In recent years, electronic cameras using a CCD image sensor, a CMOS image sensor, or the like have become widespread. However, in the process of manufacturing these image sensors, defective pixels for which normal output cannot be obtained may occur. In an image shot with an image pickup device having such defective pixels, the defective pixel portion appears as fixed noise and appears in the shot image.

In order to improve such image quality degradation, a method of calculating a predetermined correction amount to a defective pixel value whose output has been reduced to bring it close to a normal pixel value, or a pixel value of a defective pixel of an image sensor is set to a pixel of its surrounding pixels An image processing technique for interpolating from values is considered (see, for example, Patent Document 1).
JP 2005-175547 A

  However, the conventional method of simply calculating a predetermined correction amount on the pixel value of the defective pixel to obtain the corrected pixel value of the defective pixel has a slight correction error compared to the normal pixel value around the defective pixel. Therefore, when a defective pixel is corrected with a flat image with few high-frequency components, there arises a problem that the defective pixel after correction becomes conspicuous. In particular, the above-mentioned problem is remarkable when correcting a line flaw in which defective pixels are continuous.

  In addition, the conventional method of interpolating from the pixel values of the peripheral pixels is that the corrected pixel value is significantly different from the original pixel value in the case of an image with many high-frequency components in which the pixel values of the defective pixel and the peripheral pixel are significantly different. There is a risk.

  As described above, the conventional image processing technique for correcting the pixel value of the defective pixel cannot always obtain an appropriate correction pixel value at a portion where the image structure is flat or a portion having a high frequency component, and the image quality of the captured image is deteriorated. There was a problem of being.

  An object of the present invention is to obtain an appropriate corrected pixel value of a defective pixel in a portion where the image structure is flat or a portion having a high frequency component even when the image pickup device has a defective pixel, and the image quality of the photographed image is deteriorated. An object of the present invention is to provide an image processing apparatus, an electronic camera, and an allowable value setting method.

  An image processing apparatus according to a first invention calculates a first correction amount for a pixel value of a defective pixel of an image sensor and calculates a first correction pixel value; A correction pixel value allowable range setting unit that sets a correction pixel value allowable range including the first correction pixel value and a pixel value of pixels around the defective pixel are subjected to an interpolation calculation to calculate a second correction pixel value A second correction pixel value calculation unit that compares the second correction pixel value with the correction pixel value allowable range to calculate a third correction pixel value; and And a correction value substitution unit that uses a third correction pixel value as a pixel value of the defective pixel.

  According to a second invention, in the first invention, the upper limit value and the lower limit value of the correction pixel value allowable range set by the correction pixel value allowable range setting unit are the first correction pixel values for a plurality of shooting conditions in advance. It is set based on the result of evaluating the correction error with respect to the normal pixel value.

  According to a third invention, in the first or second invention, the third correction pixel value calculation unit is configured to determine the upper limit when the second correction pixel value is larger than the upper limit value of the correction pixel value allowable range. A value is calculated as the third correction pixel value, and when the second correction pixel value is smaller than the lower limit value of the correction pixel value allowable range, the lower limit value is calculated as the third correction pixel value. When the second correction pixel value is within the correction pixel allowable range, the second correction pixel value is calculated as the third correction pixel value.

  In a fourth aspect based on any one of the first to third aspects, the second correction pixel value calculation unit calculates similarity of pixel values based on pixel values of pixels around the defective pixel. The second correction pixel value is calculated by performing an interpolation operation on the defective pixel based on a pixel value that is determined and adjacent in a direction in which the similarity is high.

  According to a fifth aspect of the present invention, there is provided an electronic camera according to the above invention, wherein the image processing unit captures a subject image using an image sensor having a defective pixel and outputs the captured image to the image processing unit, and the image processing unit performs processing. And a storage unit for storing the captured image.

  An allowable value setting method according to a sixth aspect of the present invention is an allowable value setting method used in the image processing apparatus according to the first aspect of the present invention, wherein the first corrected pixel value with respect to the normal pixel value for a plurality of photographing conditions in advance. The degree of correction error is evaluated, and the correction pixel value allowable range is determined based on the evaluation result.

  According to the present invention, when a corrected pixel value of a defective pixel is obtained, a portion where the image structure is flat or a portion having a high frequency component is provided by evaluating the similarity of neighboring pixels or by providing an allowable value of the calculated corrected pixel value. Regardless, an appropriate corrected pixel value of the defective pixel can be obtained.

  Embodiments relating to an image processing apparatus, an electronic camera, and an allowable value setting method according to the present invention will be described below.

(First embodiment)
The electronic camera 101 according to the first embodiment will be described in detail with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration of an electronic camera 101 according to the present embodiment. The electronic camera 101 includes a lens 102, an image sensor 103, an analog front end (AFE) 104, an A / D converter 105, a RAW memory 106, an image processing circuit 107, a system bus 108, a memory 109, and the like. , A flash memory 110, a memory card IF 111, a memory card 112, an operation panel 113, a CPU (central processing unit) 114, a display circuit 115, a liquid crystal monitor 116, and a timing generator (TG) 117. Is done.

  Note that the electronic camera 101 according to the first embodiment includes the function of an image processing apparatus using the allowable value setting method according to the present invention. In particular, the image sensor 103 is a Bayer array image sensor having defective pixels such as line scratches. Further, processing for correcting the pixel values of these defective pixels is performed by software of the image processing circuit 107 and the CPU 114. Detailed correction processing will be described later.

  In FIG. 1, the subject light incident from the lens 102 forms an image on the light receiving surface of the image sensor 103. The image sensor 103 converts the formed subject light into an electrical signal and outputs it to the AFE 104.

  The AFE 104 performs noise removal and level adjustment of the analog electrical signal photoelectrically converted by the image sensor 103 and outputs the result to the A / D converter 105.

  The A / D converter 105 converts an analog electric signal input from the AFE 104 into digital image data. The digital image data converted by the A / D conversion unit 105 is temporarily stored in the RAW memory 106. Note that the raw image data read from the image sensor 103 is stored in the RAW memory 106. In the present embodiment, color image data obtained via a Bayer color filter is stored.

  The image processing circuit 107 reads image data from the RAW memory 106 via the system bus 108, performs image processing such as white balance processing and color interpolation processing, and stores the image data in the memory 109. Further, the image processing circuit 107 performs image compression processing according to the JPEG standard or the like on the image data stored in the memory 109 and saves it in the flash memory 110 or the memory card 112 that can be detached via the memory card I / F 111. .

  The CPU 114 operates in accordance with a program stored in the CPU 114 in advance, and controls the entire electronic camera 101 in accordance with operation information input from the operation panel 113 by the photographer. For example, the image processing contents (white balance processing, color interpolation processing, JPEG compression processing, defective pixel correction processing, etc.) are instructed to the image processing circuit 107 via the system bus 108 or displayed on the liquid crystal monitor 116 on the display circuit 115. Output images and text. When the release button on the operation panel 113 is pressed, the TG 117 is instructed to read out the subject image captured by the image sensor 103.

  The TG 117 gives an operation timing signal to the image sensor 103, the AFE 104, and the A / D converter 105 in accordance with a command from the CPU 114. The image sensor 103, the AFE 104, and the A / D converter 105 read an image signal from the image sensor 103, a noise removal operation at the AFE 104, and an analog signal at the A / D converter 105 according to the timing signal from the TG 117. Is converted into digital data.

  Here, the image data stored in the RAW memory 106 will be described in detail. As described above, the image sensor 103 captures an image through a color filter with a Bayer array. FIG. 2A illustrates a state of image data captured by the image sensor 103 and stored in the RAW memory 106. 2A is drawn as 6 pixels in the horizontal direction and 5 pixels in the vertical direction for easy understanding, but the actual image sensor 103 has millions of pixels.

  The Bayer color filter, for example, has green (G) pixels and blue (B) pixels alternately arranged in the Y1 row, and red (R) pixels and green pixels alternately in the Y2 row. Is arranged. Further, in the Y3rd row and the Y5th row, green pixels and blue pixels having the same arrangement as the Y1th row are alternately arranged. In the Y4th row, red pixels and green pixels having the same arrangement as in the Y2th row are alternately arranged.

  Here, in FIG. 2B, it is assumed that the pixel G8 in the Y3th row and the X3th column is a defective pixel, and the case where the pixel value of the defective pixel is obtained by the conventional technique will be described. For easy understanding, a very dark subject image is projected on each pixel in the Y1 and Y2 rows and in the X1 and X2 columns, and from the X3 column among the Y3 to Y5 pixels. A case where a very bright subject image is projected on the 12 pixels in the X5 column will be described. That is, the defective pixel G8 originally has a bright pixel value, and the pixels B5, G11, and R5 around the defective pixel G8 also have bright pixel values. The pixel G5, the pixel R2, the pixel G4, the pixel B4, and the pixel G10 have dark pixel values.

  FIG. 5 is a block diagram showing processing of the image processing apparatus 200 that corrects the defective pixel G8 of FIG. 5 is processed by software by the image processing circuit 107 and the CPU 114 in FIG. 1, for example.

  Hereinafter, processing from a state in which image data with defective pixels temporarily stored in the RAW memory 106 is input to the correction pixel value calculation unit S201 and the correction value substitution unit S202 will be described in order.

  (Correction Pixel Value Calculation Unit S201) In FIG. 2B, the correction pixel value of the defective pixel G8 is calculated. In a conventional method for calculating a correction pixel value, the pixel value of a defective pixel is measured, and the correction value is obtained by comparing the measured pixel value of the defective pixel with a normal pixel value. For example, since the pixel value of the defective pixel appears as a 20% output decrease or 50% output decrease of the pixel value of the normal pixel, the multiplication value that corrects this output decrease to 100% is set as the correction amount. In this case, the correction amount for a defective pixel with a 20% output decrease is five times, and the correction amount for a defective pixel with a 50% output decrease is doubled. In this way, the correction amount for the pixel value of the defective pixel is stored in the memory in advance, and the correction pixel value is calculated by multiplying the pixel value of the defective pixel of the image input at the time of shooting by this correction amount. In this method, a slight correction error occurs with respect to surrounding normal pixel values, and there is a problem that defective pixels stand out in a region where the image structure is flat.

  Alternatively, as another conventional technique, for example, there is a method in which an average value of surrounding pixels is used as a correction pixel value. In this case, the calculation is performed as follows. Since the defective pixel G8 is a green pixel, the average value of the surrounding four green pixels G4, G5, G10, and G11 is obtained and used as the corrected pixel value of the defective pixel G8.

  (Correction value substitution unit S202) The image obtained by replacing the correction pixel value obtained in step S201 as the pixel value of the defective pixel G8 is stored in the memory 109 as a corrected image. However, the former prior art described in the correction pixel value calculation unit S201 has a problem that a slight correction error occurs with respect to surrounding normal pixel values, and defective pixels are conspicuous in a region where the image structure is flat. In the latter prior art, for example, in the case of FIG. 2B, the pixel G4, the pixel G5, and the pixel G10 are dark pixels, and only the pixel G11 is a bright pixel. Therefore, the defective pixel G8 should originally have a bright pixel value. However, the corrected pixel value obtained by this method is a considerably dark pixel value. Therefore, the corrected image stored in the memory 109 is replaced with a very dark pixel value different from the original defective pixel G8, as shown in FIG. 6, and the image structure has a large high-frequency component. Then problems arise.

  As described above, the correction of the defective pixel according to the conventional technique shown in FIG. 5 has a problem that the original pixel value is not necessarily obtained and is corrected erroneously. In the above description, the case where the defective pixel is a point defect pixel including only the pixel G8 has been described for easy understanding. However, when the defective pixel is a continuous line defect, a correction error is particularly noticeable in a portion where the image structure is flat. End up. In general, the characteristics of line scratches are such that the pixel output on line scratches is lower than that on normal pixels, but there is an output that is roughly proportional to the amount of exposure, and the rate of output drop for each pixel on the same line scratch is approximately Often the same. For this reason, in the case of the method of correcting the line flaw region with a fixed correction amount, an error with the peripheral pixels in the line flaw region is particularly noticeable.

  On the other hand, the electronic camera 101 according to the present embodiment can appropriately correct the pixel value of the defective pixel regardless of whether the image structure is a flat region or a region having a large high-frequency component. Thus, an image with little correction error can be obtained. Next, processing of the electronic camera 101 according to the first embodiment using the image processing apparatus and the allowable value setting method according to the present invention will be described in detail with reference to FIG.

  FIG. 3 is a block diagram illustrating processing of the image processing apparatus 100 of the electronic camera 101 according to the present embodiment. Note that the image processing apparatus 100 in FIG. 3 is processed by software by the image processing circuit 107 and the CPU 114 in FIG. In FIG. 3, image data with defective pixels temporarily stored in the RAW memory 106 includes a first correction pixel value calculation unit S101, a second correction pixel value calculation unit S103, and a correction value substitution unit S105. And have been entered. Hereinafter, correction processing from this state will be described in order.

  (First Correction Pixel Value Calculation Unit S101) The first correction pixel value calculation unit S101 calculates a first correction pixel value for the defective pixel of the input image. The position of the defective pixel is assumed to be known in the previous process and is stored in advance in the flash memory 110 or the like.

  For example, in FIG. 2B, when calculating the first correction pixel value of the defective pixel G8, the pixel value of the defective pixel G8 is multiplied by a predetermined correction amount set in advance to correct the defective pixel G8. This is the first corrected pixel value. The predetermined correction amount is obtained, for example, in an adjustment process before the electronic camera 101 is manufactured and shipped at the factory, and is stored in the flash memory 110 or the like.

  Here, how to determine the predetermined correction amount will be described. First, the pixel value in the defective pixel is measured for each of a plurality of images shot under various shooting conditions such as the exposure amount. In the case of a line flaw, the pixel values at a plurality of defective pixels on the line flaw are measured. Next, using the correction amount to be obtained as an unknown variable, for example, using a mathematical method such as a least square method, the correction error between the correction result obtained by multiplying the pixel value obtained by measuring the correction amount and the normal pixel value is minimized. A correction amount that is an unknown variable is obtained.

  The correction amount obtained in this way is stored in the flash memory 110 as a predetermined correction amount in association with each defective pixel position (for each address of the image sensor 103). In the case of continuous pixels such as line scratches, one value may be stored in the flash memory 110 as a predetermined correction amount for each line scratch area (for each address range of the image sensor 104).

  As described above, the first correction pixel value calculation unit S101 obtains the first correction pixel value by multiplying the pixel value of the defective pixel by the predetermined correction amount stored in the flash memory 110 at the time of shipment from the factory. Deliver to processing step.

  (Correction pixel value allowable range setting unit S102) The correction pixel value allowable range setting unit S102 sets the allowable range of correction pixel values based on the first correction pixel value calculated by the first correction pixel value calculation unit S101. An upper limit value and a lower limit value to be expressed are calculated.

For example, the upper limit value and the lower limit value of the allowable range of the correction pixel value can be calculated by the following equation using a predetermined allowable rate set to correct the pixel value of the defective pixel.
Upper limit value = first correction pixel value × (1 + allowance rate) (Expression 1)
Lower limit value = first correction pixel value × (1−allowance rate) (Expression 2)
Here, the allowable rate is obtained in an adjustment process performed before the electronic camera 101 is manufactured at the factory and is stored in the flash memory 110 or the like. For example, as a method for obtaining the tolerance, first, pixel values at a plurality of pixel positions on a line scratch are measured for a plurality of images photographed under various photographing conditions with different exposure amounts and the like. In the case of a point defect pixel, the pixel value at the pixel position is measured. Next, a ratio between a correction value obtained by correcting the measured pixel value with the first correction pixel value calculated by the first correction pixel value calculation unit S101 and a normal pixel value is obtained. For example, when the correction value corrected with the first correction pixel value is 200, which is a pixel value of 256 gradations, and the normal pixel value is 220, the ratio is 200 / 220≈0.91. The ratio may be set to 220/200 = 1.1 by reversing the denominator and the numerator, and any ratio may be used. The ratio is 1 only when the correction value is equal to the normal pixel value. In the present embodiment, the maximum value of the error in which the ratio obtained in this way deviates from 1 is set as the allowable rate. Note that it is not always necessary to set the maximum value of the error whose ratio deviates from 1 as the allowable rate, and a value obtained by subtracting a predetermined ratio such as 10% of the maximum value or 1/2 of the maximum value may be set as the allowable rate. . The obtained allowable rate is stored in the flash memory 110 as a predetermined allowable rate corresponding to each defective pixel position (for each address of the image sensor 103). In the case of a line flaw, one value may be stored in the flash memory 110 as a predetermined allowable rate for each line flaw region (for each address range of the image sensor 104).

  For example, in FIG. 2B, when the first correction pixel value of the defective pixel G8 is 200 and the allowable rate is 0.2, the upper limit value is 240 from (Expression 1), and from (Expression 2). The lower limit value is 160.

  As described above, the correction pixel value allowable range setting unit S102 sets the upper limit value and the lower limit value for each defective pixel or each line scratch using the predetermined allowable rate stored in the flash memory 110 at the time of factory shipment (Formula 1). And it calculates by (Formula 2), and hands it over to the next processing step. The above processing is the allowable value setting method according to the present invention.

  (Second correction pixel value calculation unit S103) The second correction pixel value calculation unit S103 calculates a pixel value at the correction target pixel position by using a pixel around the correction target pixel by a predetermined interpolation method, and calculates this. The second corrected pixel value is output to the next processing step.

  Here, the predetermined interpolation method is, for example, the absolute value of the difference between two same color pixels adjacent in the horizontal direction of the interpolation target pixel and the absolute value of the difference between two same color pixels adjacent in the vertical direction of the interpolation target pixel. And the absolute value of the difference between the two same color pixels adjacent to the interpolation target pixel in the oblique 45 degree direction and the absolute value of the difference between the two same color pixels adjacent to the interpolation target pixel in the oblique 135 degree direction. The direction in which the absolute value of the difference is the smallest is determined as the “direction with high similarity”. Then, an average value of pixels of the same color that are adjacent to each other in the “high similarity direction” is obtained as a second corrected pixel value.

  For example, in the case of FIG. 2B, the two pixels of the same color that are adjacent in the horizontal direction of the defective pixel G8, which is the correction target pixel, are the pixel G7 and the pixel G9, and thus | pixel value of G7−pixel value of G9 | Similarly, since two pixels of the same color adjacent to the defective pixel G8 in the vertical direction are the pixel G2 and the pixel G14, the pixel value | G2−the pixel value | G14 is obtained. Also, since the two pixels of the same color adjacent to the defective pixel G8 in the oblique 45-degree direction are the pixel G5 and the pixel G10, | the pixel value of G5−the pixel value of G10 | is obtained. Further, since the two pixels having the same color adjacent to the defective pixel G8 in the diagonal direction of 135 degrees are the pixel G4 and the pixel G11, the pixel value | G4−the pixel value | G11 is obtained. Then, the direction in which the absolute value of these differences is minimized is determined. In the above case, since the combination of two pixels G5 and G10 of the same color adjacent to the defective pixel G8 in the oblique 45 degree direction is a dark pixel value, the absolute value of the difference is small. Since the combination of the dark pixel value and the bright pixel value is also the absolute value of the difference is larger than that in the case of 45 degrees oblique direction. Therefore, it is understood that the “direction with high similarity” is a 45-degree oblique direction, and an average value of the pixels G5 and G10 of two same-color pixels adjacent to the defective pixel G8 in the oblique 45-degree direction is obtained. The second correction pixel value is used. For example, when the pixel value of 256 gradations of the pixel G5 is 24 and the pixel G10 is 22, the average value of the pixel G5 and the pixel G10 is 23, and the second correction pixel value is 23 and output to the next processing step. To do.

  (Correction pixel value allowable range setting unit S104) The correction pixel value range restriction unit (third correction pixel value calculation unit) S104 includes the second correction pixel value calculated by the second correction pixel value calculation unit S103, and The corrected pixel value allowable range calculated in the corrected pixel value allowable range setting unit S102 is compared to calculate and output a third corrected pixel value.

  Specifically, when the second correction pixel value is larger than the upper limit value of the correction pixel value allowable range, the upper limit value is output as the third correction pixel value, and when the second correction pixel value is smaller than the lower limit value, the lower limit value is set. 3 is output as the third correction pixel value, and when it is larger than the lower limit value and smaller than the upper limit value, the second correction pixel value is output as it is as the third correction pixel value.

  For example, in FIG. 2B, the upper limit value of the defective pixel G8 calculated by the correction pixel value allowable range setting unit S102 is 240, the lower limit value is 160, and the defective pixel G8 is calculated by the second correction pixel value calculation unit S103. When the second correction pixel value is 23, it is below the lower limit value 160, so the lower limit value 160 is output as the third correction pixel value to the next step.

  (Correction Value Substitution Unit S105) The correction value substitution unit S105 replaces the third correction pixel value output from the correction pixel value range restriction unit S104 with the pixel value of the defective pixel. In the case of the above example, since the defective pixel G8 is corrected to a bright pixel value of the lower limit value 160, as shown in FIG. 4, it becomes close to the original brightness of the pixel G8. In the conventional case, as described with reference to FIG. 6, the defective pixel G8 may be erroneously corrected to a dark pixel value, but in this embodiment, it can be corrected to a value close to the original pixel value. . Thereafter, an image obtained by performing the same processing on all defective pixels and line scratches is stored in the memory 109 or the memory card 112 as a captured image.

  As described above, the electronic camera 101 according to the present embodiment uses the image sensor 103 having point defect pixels and line scratches. However, the image processing apparatus 100 including the above-described processing steps S101 to S105 causes the defective pixels to be detected. The pixel value can be appropriately corrected, and a high-quality image can be taken.

  Hereinafter, features and effects of the present embodiment will be described. In the image processing apparatus 100 shown in FIG. 3 of the electronic camera 101 according to the present embodiment, first, the first correction pixel value calculation unit multiplies the correction target pixel value, which is a defective pixel, by a predetermined correction amount. Thus, the first correction pixel value is calculated. However, since the first correction pixel value includes a slight error, it is not desirable to directly adopt the value as the correction value as in the prior art. It can be expected that the first corrected pixel value includes a true pixel value within the vicinity thereof.

  Therefore, the correction pixel value allowable range setting unit S102 of the present embodiment sets a range (upper limit value and lower limit value) of a predetermined allowable width including the first correction pixel value as the correction pixel value allowable range. In particular, the correction pixel value allowable range is the maximum error and the upper limit value so that the true pixel value is included in the correction pixel value allowable range even when the error due to the first correction pixel value is maximum. Set the lower limit.

  Further, the correction pixel value range limiting unit S104 outputs the third correction pixel value by limiting the second correction pixel value within the correction pixel value allowable range, but the second correction pixel value is true. If the pixel value is close to this pixel value, the value is included in the correction pixel value allowable range, and the value is output as the third correction pixel value.

  On the other hand, if the second correction pixel value deviates greatly from the true pixel value, the pixel value falls outside the correction pixel value allowable range, so that the second correction pixel value falls within the correction pixel value allowable range. Output as a third corrected pixel value. As a result, the third correction pixel value is always closer to the true pixel value than the second correction pixel value, and correction errors can be reduced.

  As described above, since the image processing apparatus 100 of the electronic camera 101 according to the embodiment uses the correction value obtained by the second correction pixel value calculation unit S103 as it is in a place where the image structure is flat, an appropriate correction result is obtained. can get. On the other hand, when the image structure has a large high-frequency component, the correction pixel value range limiting unit S104 corrects the correction value at a location where the correction value obtained by the second correction pixel value calculation unit S103 is inappropriate. Since the pixel value is limited within the allowable range, an almost appropriate correction result close to the true pixel value can be obtained.

  In the first correction pixel value calculation unit S101 of the present embodiment, the first correction pixel value is calculated by multiplying the correction target pixel value having a defective pixel by a predetermined correction amount. Processing is not limited to multiplication. For example, instead of multiplying the correction amount, the first correction pixel value may be calculated by performing addition / subtraction or other arithmetic processing.

  In the correction pixel value allowable range setting unit S102 of this embodiment, the upper limit value and the lower limit value of the correction pixel value allowable range are determined by multiplying the first correction pixel value by a predetermined allowable rate. However, the arithmetic processing is not limited to multiplication. For example, the correction pixel value allowable range may be determined by performing addition / subtraction or other operations on the first correction pixel value.

  In addition, in the correction pixel value range limiting unit S104 of the present embodiment, when the second correction pixel value is limited, the upper limit value or the lower limit value of the correction pixel value allowable range is output. It is not limited to. For example, a value slightly smaller than the upper limit value or a value slightly larger than the lower limit value may be output.

  In the present embodiment, the case of correcting point defect pixels and line flaws has been described. However, the present invention is not limited to this, and can be applied to pixel defect correction of an arbitrary shape.

1 is a block diagram of an electronic camera 101 according to a first embodiment. It is explanatory drawing which shows a pixel arrangement | sequence. It is a block diagram which shows the process of the image processing apparatus 100 of the electronic camera 101 which concerns on 1st Embodiment. It is explanatory drawing which shows the example of a correction | amendment in 1st Embodiment. It is a block diagram which shows the process of the conventional image processing apparatus. It is explanatory drawing which shows the example of a conventional correction | amendment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Image processing apparatus 101 ... Electronic camera 102 ... Lens 103 ... Imaging device 106 ... Raw memory 107 ... Image processing circuit 109 ... Memory 110 ... Flash memory 112- ..Memory card 114 ... CPU
116: Liquid crystal monitor S101: First correction pixel value calculation unit S102: Correction pixel value allowable range setting unit S103: Second correction pixel value calculation unit S104: Correction pixel value range limitation Part S105... Correction value substitution part

Claims (6)

A first correction pixel value calculation unit that calculates a first correction pixel value by calculating a predetermined correction amount with respect to the pixel value of the defective pixel of the image sensor;
A correction pixel value allowable range setting unit for setting a correction pixel value allowable range including the first correction pixel value;
A second correction pixel value calculation unit that performs an interpolation operation using pixel values of pixels around the defective pixel and calculates a second correction pixel value;
A third correction pixel value calculating unit that calculates a third correction pixel value by comparing the second correction pixel value and the correction pixel value allowable range;
An image processing apparatus comprising: a correction value substitution unit that uses the third correction pixel value as a pixel value of the defective pixel. The image processing apparatus according to claim 1.
The upper limit value and the lower limit value of the correction pixel value allowable range set by the correction pixel value allowable range setting unit are obtained as a result of evaluating a correction error with respect to the normal pixel value of the first correction pixel value in advance for a plurality of shooting conditions. An image processing apparatus that is set based on the above. The image processing apparatus according to claim 1 or 2,
The third correction pixel value calculation unit calculates the upper limit value as the third correction pixel value when the second correction pixel value is larger than the upper limit value of the correction pixel value allowable range, When the second correction pixel value is smaller than the lower limit value of the correction pixel value allowable range, the lower limit value is calculated as the third correction pixel value, and the second correction pixel value falls within the correction pixel allowable range. In some cases, the second correction pixel value is calculated as the third correction pixel value. The image processing apparatus according to any one of claims 1 to 3,
The second correction pixel value calculation unit determines similarity of pixel values based on pixel values of pixels around the defective pixel, and determines the defect based on pixel values adjacent in a direction in which the similarity is high. An image processing apparatus that performs an interpolation operation on a pixel to calculate the second corrected pixel value.   5. The image processing apparatus according to claim 1, wherein the image processing unit captures a subject image using an image sensor having a defective pixel and outputs the captured image to the image processing unit, and the image processing. An electronic camera, further comprising a storage unit that stores a captured image processed by the unit. An allowable value setting method used in the image processing apparatus according to claim 1,
An allowable value setting method characterized by evaluating a degree of correction error of the first correction pixel value with respect to a normal pixel value in advance for a plurality of photographing conditions and determining the correction pixel value allowable range based on the evaluation result .

Images