JP2007267072A - Electronic camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an electronic camera with which quality of photographed image data is improved. <P>SOLUTION: The electronic camera 10 includes a CPU 26 and a CCD imager 14 has an imaging surface 14f on which a plurality of pixels are two-dimensionally arranged. Raw image data expressing a field captured by the imaging surface 14f is created by a CDS/AGC/AD circuit 16. Luminance values of eight peripheral pixels existing in the periphery of a defective pixel among the plurality of pixels are detected by the CPU 26 from the raw image data created by the CDS/AGC/AD circuit 16. The CPU 26 discriminates whether or not lines over the defective pixel is drawn on a partial image area expressed by the defective pixel and the eight peripheral pixels based on the plurality of detected luminance values and compensate a luminance value of the defective pixel in different forms according to discrimination results. Exact defective pixel compensation is attained. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an electronic camera, and more particularly to an electronic camera that corrects defective pixels included in an image sensor, for example.

An example of a conventional device of this type is disclosed in Patent Document 1. According to this prior art, defective pixels of the image sensor are detected based on black image data generated by the image sensor by exposure with the mechanical shutter closed. The coordinate data of the detected defective pixel is temporarily stored in the memory. The defective pixel of the image data representing the object scene is corrected by the median filter process with reference to the coordinate data stored in the memory.
JP 2004-23331 A [H04N 5/335]

  However, in the related art, when a scene with high contrast is photographed, so-called pixel collapse occurs in the pixel data of the defective pixel, and the quality of the photographed image data is degraded.

  Therefore, a main object of the present invention is to provide an electronic camera capable of improving the quality of captured image data.

  An electronic camera (10) according to the invention of claim 1 includes an imaging means (14) having an imaging surface (14f) in which a plurality of pixels are two-dimensionally arranged, and image data representing an object scene captured by the imaging surface. Creating means (16) for creating, detecting means (S21) for detecting data values of a plurality of peripheral pixels existing around the defective pixel among the plurality of pixels from the image data created by the creating means, the defective pixel and the plurality of pixels Discriminating means for discriminating whether or not a line straddling the defective pixel is drawn on the partial image represented by the peripheral pixels based on a plurality of data values detected by the detecting means (S47, S49, S51, S53) And correction means (S55, S57, S33) for correcting the data value of the defective pixel in different manners according to the discrimination result of the discrimination means.

  The imaging means has an imaging surface on which a plurality of pixels are arranged two-dimensionally. Image data representing the object scene captured on the imaging surface is created by the creating means. Data values of a plurality of peripheral pixels existing around the defective pixel among the plurality of pixels are detected by the detection unit from the image data generated by the generation unit. The discriminating unit discriminates whether or not a line straddling the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels based on the plurality of data values detected by the detecting unit. The correction unit corrects the data value of the defective pixel in a different manner according to the determination result of the determination unit.

  The correction mode of the data value of the defective pixel differs depending on whether or not a line straddling the defective pixel is drawn. Thus, accurate defective pixel correction is realized.

  The electronic camera according to the invention of claim 2 is dependent on claim 1 and further comprises classification means (S45) for classifying the plurality of peripheral pixels into low-intensity pixels and high-intensity pixels. First pattern discriminating means (S51) for discriminating whether or not the predetermined first pattern is met, second pattern discriminating means (S53) for discriminating whether or not a plurality of peripheral pixels follow the predetermined second pattern, and Activating means (S47, S49) for activating one of the first pattern discriminating means and the second pattern discriminating means based on the classification result of the classification means is included.

  According to the second aspect of the present invention, when a line straddling the defective pixel is not drawn on the partial image, it is determined by the first pattern determining means and the second pattern determining means. As a result, defective pixels are quickly corrected.

  An electronic camera according to a third aspect of the invention is dependent on the second aspect, and when the determination result of the first pattern determination means is negative, the correction means refers only to the low-luminance pixels among the low-luminance pixels and the high-luminance pixels. When the determination results of the first correction unit and the second pattern determination unit are negative, the correction process is executed with reference to only the high luminance pixels of the low luminance pixels and the high luminance pixels. Including correction means.

  According to the invention of claim 3, when a plurality of peripheral pixels do not follow the predetermined first pattern, it is determined that a low-luminance line straddling the defective pixel is drawn on the partial image. That is, it is determined that the brightness of the defective pixel indicates brightness approximate to that of the low luminance pixel. The defective pixel is corrected by the first correction unit that refers only to the low luminance pixel among the low luminance pixel and the high luminance pixel. In addition, when the plurality of peripheral pixels do not follow the predetermined second pattern, it is determined that a high-brightness line across the defective pixel is drawn on the partial image. That is, it is determined that the brightness of the defective pixel indicates brightness approximate to that of the low luminance pixel. The defective pixel is corrected by the second correcting unit that refers to only the high luminance pixel among the low luminance pixel and the high luminance pixel. As a result, an accurate correction process in consideration of the relationship between the line and the background is realized.

  An electronic camera according to a fourth aspect of the invention is dependent on the third aspect, and the correction means has a low luminance when one of the determination result of the first pattern determination means and the determination result of the second pattern determination means is affirmative. Third correction means for executing correction processing with reference to both the pixel and the high luminance pixel is further included. As a result, the defective pixel is accurately corrected.

  The electronic camera according to the invention of claim 5 is dependent on any one of claims 1 to 4 and further comprises holding means (52) for holding pixel information for identifying defective pixels, and the detection means is held by the holding means. A detection process is executed based on the pixel information.

  The electronic camera according to the invention of claim 6 is dependent on claim 5, and the pixel information of the defective pixel held by the holding means is position information indicating the position where the defective pixel appears and brightness information indicating the brightness of the defective pixel. Including.

  An electronic camera according to a seventh aspect of the invention is dependent on the first or sixth aspect, further comprising a color filter having a plurality of color elements respectively covering a plurality of pixels, each of the plurality of color elements being one of a plurality of colors. And the color of the color filter covering the surrounding pixels matches the color of the color filter covering the defective pixel.

  An image processing program according to an invention of claim 8 creates image data representing an object scene captured by an imaging means (14) having an imaging surface (14f) in which a plurality of pixels are two-dimensionally arranged. The processor (26) of the electronic camera (10) having the creation means (16) detects data values of a plurality of peripheral pixels existing around the defective pixel among the plurality of pixels from the image data created by the creation means. Detection step (S21), based on the plurality of data values detected by the detection step, whether or not a line across the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels Image processing program for executing a discrimination step (S47, S49, S51, S53) and a correction step (S55, S57, S33) for correcting the data value of the defective pixel in a different manner according to the discrimination result of the discrimination step It is.

  An image processing method according to a seventh aspect of the present invention provides an image pickup means (14) having an image pickup surface (14f) in which a plurality of pixels are two-dimensionally arranged, and image data representing an object scene captured on the image pickup surface. An image processing method for an electronic camera (10) comprising a creation means (16), wherein data values of a plurality of peripheral pixels existing around a defective pixel among a plurality of pixels are detected from image data created by the creation means Detecting step (S21), determining whether a line across the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels based on the plurality of data values detected by the detecting step An image processing method comprising a determination step (S47, S49, S51, S53) and a correction step (S55, S57, S33) for correcting the data value of the defective pixel in a different manner according to the determination result of the determination step .

  In the seventh and eighth aspects, as in the first aspect, the correction mode of the data value of the defective pixel differs depending on whether a line straddling the defective pixel is drawn. Accurate defective pixel correction is realized.

  According to this invention, the correction mode of the data value of the defective pixel differs depending on whether or not a line straddling the defective pixel is drawn. Thus, accurate defective pixel correction is realized.

  The above object, other objects, features and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

  Referring to FIG. 1, a digital camera (electronic camera) 10 of this embodiment includes an optical lens 12. The optical image of the object scene is irradiated through the optical lens 12 onto the light receiving surface of the CCD imager 14, that is, the imaging surface 14f.

  The imaging surface 14f is covered with a primary color Bayer array color filter (see FIG. 2). The plurality of filter elements constituting the color filter respectively correspond to the plurality of light receiving elements forming the imaging surface 14f. Therefore, the amount of charge generated in each light receiving element reflects one of the light amounts of R (Red), G (Green), and B (Blue). The light receiving element is defined as “pixel”.

  The plurality of pixels forming the imaging surface 14f are divided into normal pixels and defective pixels. The defective pixel includes a highly sensitive defective pixel that outputs a maximum amount of charge compared to a normal pixel and a highly sensitive defective pixel that outputs a minimum amount of charge compared to a normal pixel.

  When the power is turned on, a defective pixel information reading process and a through image process are executed. Specifically, the CPU 26 first reads out defective pixel information stored in the flash memory 52 via the bus B1. Here, the defective pixel information includes position information Adr of a plurality of defective pixels included in the CCD imager 14 and is stored in a memory 26 m formed in the CPU 26. The position information Adr represents the position of the defective pixel.

  Subsequently, the CPU 26 connects the switches SW1 and SW2 to the terminals T1 and T3, respectively, instructs the timing generator (TG) / signal generator (SG) 18 to repeat the pre-exposure and thinning-out reading, and writes the memory control circuit 38. The destination is set in the through image area 40a of the SDRAM 40.

  Referring to FIG. 3, SDRAM 40 has a through image area 40a, a raw image area 40b, a work area 40c, and a compressed image area 40d. The through image area 40a is used after the power is turned on, the raw image area 38b is used after a recording operation, that is, a full pressing operation by the shutter button 54, and the work area 40c is used during pixel correction processing (described later). The compressed image area 40d is used during JPEG compression processing (described later).

  Returning to FIG. 1, the TG / SG 18 generates a vertical synchronization signal Vsync every 1/30 seconds, and provides a plurality of timing signals synchronized therewith to the CCD imager 14 and the CDS / AGC / AD circuit 16. The imaging surface 14f is subjected to pre-exposure every time the vertical synchronization signal Vsync is generated, and the charge generated on the imaging surface 14f is subjected to thinning readout by raster scanning. The low-resolution raw image signal formed by the read charges has a frame rate of 30 fps.

  The CDS / AGC / AD circuit 16 subjects the raw image signal output from the CCD imager 14 to a series of processes of correlated double sampling, gain adjustment and A / D conversion, and outputs raw image data that is a digital signal. The raw image data output from the CDS / AGC / AD circuit 16 is given to the signal processing circuit 20 via the switches SW1 and SW2.

  The signal processing circuit 20 performs processing such as white balance adjustment, color separation, and YUV conversion on the given raw image data, and YUV format image data generated thereby is subjected to memory control via the buffer circuit 22 and the bus B1. The data is supplied to the circuit 38 and written into the through image area 40 a of the SDRAM 40 by the memory control circuit 38.

  The video encoder 44 reads out the image data thus stored in the SDRAM 40 through the memory control circuit 38. The read image data is supplied to the video encoder 44 through the bus B1 and the buffer circuit 42, and converted into an NTSC composite video signal. The converted composite video signal is applied to an LCD (Liquid Crystal Display) 46, and as a result, a through image representing the object scene is reproduced on the LCD 46.

  The luminance evaluation circuit 24 evaluates the luminance of the object scene every 1/30 seconds based on the Y data among the YUV format image data generated by the signal processing circuit 20. The evaluation result, that is, the luminance evaluation value is used for the through image AE processing by the CPU 26. The CPU 26 calculates the optimum exposure time based on the luminance evaluation value, and instructs the TG / SG 18 to perform pre-exposure processing according to the calculated optimum exposure time. As a result, the brightness of the through image displayed on the LCD 46 is appropriately adjusted.

  When the shutter button 54 is half-pressed, the CPU 26 executes a recording AE process. The pre-exposure time of the CCD imager 14 is adjusted more strictly based on the above-described luminance evaluation value. The optimum exposure time set in the TG / SG 18 by the recording AE process is not changed as long as the shutter button 54 is half-pressed.

  When the shutter button 54 is fully pressed, the photographing process is executed. Specifically, the CPU 26 connects the switches SW1 and SW2 to the terminals T2 and T4, instructs the TG / SG 18 to execute main exposure and all-pixel reading according to the optimum exposure time, and controls the change of the write setting. Command circuit 38. By changing the writing setting, the raw image area 40b (see FIG. 3) is set as the writing destination.

  The TG / SG 18 performs the main exposure on the CCD imager 14 according to the optimum exposure time calculated by the recording AE process, and reads out all the generated charges in a raster scanning manner. At this time, the imaging surface 14f is scanned in an interlaced manner. A raw image signal having a high pixel number formed by all charges is output from the CCD imager 14 over a period of two fields (= 1/30 seconds * 2).

  The raw image signal of each field output from the CCD imager 14 is subjected to the same processing as described above by the CDS / AGC / AD circuit 16 and converted into raw image data. The converted raw image data is given to the buffer circuit 28 via the switch SW1, and then written to the raw image area 40b of the SDRAM 40 via the bus B1 and the memory control circuit 38.

  When raw image data is stored in the raw image area 40b, pixel correction processing is executed. Specifically, the CPU 26 first generates a raw image from the defective pixel luminance value DEFECT and the peripheral eight luminance values Laound [0] to Round [7] based on the defective pixel position information Adr registered in the table 26t. Detect from data.

  As shown in FIG. 4, each of the peripheral eight pixels is covered with a filter element having the same color as the color of the filter element covering the defective pixel. By paying attention to the same color in this way, the defective pixel is accurately corrected.

  Returning to FIG. 1, the CPU 26 then arranges the luminance values Lauround [0] to Laound [7] of the surrounding eight pixels in ascending order, and obtains the aligned luminance values Lsort [0] to Lsort [7] thus obtained. The data is stored in the work area 40c (see FIG. 3) of the SDRAM 40. It should be noted that the highest luminance value among the luminance values around [7] to [7] of the surrounding eight pixels is replaced with the aligned luminance value Lsort [0], and the luminance values from [round] to [7] of the surrounding eight pixels. The lowest luminance value is replaced with the aligned luminance value Lsort [7].

  When the aligned luminance values Lsort [0] to Lsort [7] are stored, the CPU 26 subsequently determines whether or not the partial image represented by the defective pixel and the surrounding eight pixels is a low contrast image. Specifically, it is determined whether or not the numerical value obtained by dividing the aligned luminance value Lsort [1] by the aligned luminance value Lsort [6] is equal to or greater than a constant Gap. In addition, the reliability of the determination result is improved by employing the alignment luminance value Lsort [1] and the alignment luminance value Lsort [6] in the calculation formula for the value to be compared with the constant Gap.

  If the partial image represented by the defective pixel and the surrounding eight pixels is a low-contrast image, the defective pixel is corrected by the median filter process. Specifically, the luminance value DEFECT of the defective pixel and the luminance values Laound [0] to Laound [7] of the surrounding eight pixels are arranged in descending order, and the median value of the arranged nine luminance values, that is, the lowest luminance value is set. The fifth luminance value is replaced with the luminance value DEFECT of the defective pixel. In other words, all of the luminance values around [0] to [7] of the surrounding eight pixels are referred to during the median filter processing.

  On the other hand, if the partial image represented by the defective pixel and the surrounding 8 pixels is a high-contrast image, the CPU 26 further determines the reliability of the sample used for the average value filter processing in the surrounding 8 pixels of the defective pixel. It is determined whether or not a pixel exists. If there are defective pixels in the surrounding 8 pixels of the defective pixel, the luminance values Laound [0] to Laound [7] of the surrounding 8 pixels are regarded as a sample with low reliability for the average value filter processing. Median filtering is performed.

  On the other hand, if there are no defective pixels in the surrounding 8 pixels of the defective pixel, the luminance values Laound [0] to Laound [7] of the surrounding 8 pixels are regarded as highly reliable samples for the average value filtering process. The average value filtering process is executed.

  Specifically, the CPU 26 first calculates a threshold value TH. The threshold value TH is equal to ½ of the numerical value obtained by adding the aligned luminance value Lsort [1] and the aligned luminance value Lsort [6] to each other. In addition, the defective pixel is quickly corrected by adopting the alignment luminance value Lsort [1] and the alignment luminance value Lsort [6] in the calculation formula of the threshold TH.

  Next, the CPU 26 classifies the aligned luminance values that exceed the threshold TH among the aligned luminance values Lsort [0] to Lsort [7] into high luminance pixels, and the threshold TH among the aligned luminance values Lsort [0] to Lsort [7]. The following aligned luminance values are classified into high luminance pixels.

  When the number of pixels classified as low-luminance pixels among the eight peripheral pixels is “2” or “3”, each of the pixels classified as low-luminance pixels among the eight peripheral pixels is the luminance value that the defective pixel should originally indicate. The CPU 26 determines whether or not the surrounding 8 pixels correspond to the exclusion pattern 1 (see FIGS. 5A to 5P). When the number of low luminance pixels is “2”, the number of pixels classified as high luminance pixels is “6”. When the number of low luminance pixels is “3”, the number of pixels classified as high luminance pixels is “5”. It is a piece.

  When a defective pixel exists in the center of the image area as shown in FIG. 6A, the surrounding 8 pixels correspond to the exclusion pattern 1 (see FIG. 5J), and thus the above-described median filter processing is executed. . As a result, as shown in FIG. 6B, the luminance value DEFECT of the defective pixel is replaced with any one of the luminance values classified into the low luminance pixels.

  When a defective pixel exists in the center of the image area as shown in FIG. 7A, the peripheral 8 pixels do not correspond to the exclusion pattern 1, and thus the defective pixel forms a line represented by a low luminance value. It is judged. At this time, the CPU 26 calculates the average value of the aligned luminance values classified into the low luminance pixels, and replaces the luminance value DEFECT of the defective pixel with the calculated average value. As a result, as shown in FIG. 7B, the defective pixel has a luminance value approximate to the low luminance pixel formed by the line represented by the low luminance value.

  If there are “2” or “3” pixels classified as high-intensity pixels among the eight peripheral pixels, the CPU 26 should originally indicate each pixel classified as a high-luminance pixel among the eight peripheral pixels as a defective pixel. It is determined that the brightness value approximates, and it is determined whether or not the surrounding 8 pixels correspond to the exclusion pattern 2 (see FIGS. 8A to 8P). When the number of high luminance pixels is “2”, the number of pixels classified as low luminance pixels is “6”. When the number of high luminance pixels is “3”, the number of pixels classified as low luminance pixels is “5”. It is a piece.

  When a defective pixel exists in the center of the image region as shown in FIG. 9A, the surrounding 8 pixels correspond to the exclusion pattern 2 (see FIG. 8C), and thus the above-described median filter processing is executed. . As a result, as shown in FIG. 9B, the luminance value DEFECT of the defective pixel is replaced with any one of the luminance values classified as high luminance pixels.

  When a defective pixel exists in the center of the image area as shown in FIG. 10A, the peripheral 8 pixels do not correspond to the exclusion pattern 2, and thus the defective pixel is a high luminance pixel that forms a line represented by a high luminance value. It is judged. At this time, the CPU 26 calculates an average value of the aligned luminance values classified into the high luminance pixels, and replaces the luminance value DEFECT of the defective pixel with the calculated average value. As a result, as shown in FIG. 10B, the defective pixel has a luminance value that approximates the high luminance pixel formed by the line represented by the high luminance value.

  That is, when the surrounding 8 pixels are along the excluded 1 group, it is determined that a low-luminance line across the defective pixel is drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. In other words, it is determined that the brightness of the defective pixel indicates brightness approximating that of the low luminance pixel. The CPU 26 corrects the defective pixel by referring to only the low luminance pixel among the low luminance pixel and the high luminance pixel. Further, when the surrounding 8 pixels are along the excluded 2 group, it is determined that a high-brightness line across the defective pixel is drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. In other words, it is determined that the brightness of the defective pixel indicates brightness approximating that of the high luminance pixel. The CPU 26 corrects the defective pixel by referring to only the high luminance pixel among the low luminance pixel and the high luminance pixel. As a result, an accurate correction process in consideration of the relationship between the line and the background is realized. A part of the peripheral 8 pixels is referred to when the average value filter processing is performed.

  The above-described processing is continued until the correction processing for all defective pixels is completed, and the work area 40c of the SDRAM 40 is used.

  When the pixel correction processing of the raw image data is completed, the CPU 26 gives an instruction to the signal processing circuit 20, the JPEG encoder 34, the memory control circuit 38, and the I / F circuit 48 in order to execute the recording processing. The signal processing circuit 20 reads the raw image data stored in the raw image area 40b through the memory control circuit 38 in a progressive scanning manner. The read raw image data is supplied to the signal processing circuit 20 through the bus B1, the buffer circuit 30, and the switch SW2, and is converted into image data in the YUV format. The converted image data is applied to the memory control circuit 38 via the buffer circuit 22 and the bus B1, and is written into the SDRAM 40 by the memory control circuit 38.

  When the switches SW1 and SW2 are connected to the terminals T1 and T3, respectively, the raw image data output from the CDS / AGC / AD circuit 16 is given to the signal processing circuit 20. On the other hand, when the switches SW1 and SW2 are connected to the terminals T2 and T4, the raw image data output from the CDS / AGC / AD circuit 16 is applied to the SDRAM 40 through the buffer circuit 28, the bus B1 and the memory control circuit 38. It is done.

  The JPEG encoder 34 reads out the image data stored in the raw image area 40b through the memory control circuit 38. The read image data is given to the JPEG encoder 34 through the bus B1 and the buffer circuit 32, and subjected to JPEG compression. The compressed image data generated thereby is given to the memory control circuit 38 via the buffer circuit 36 and the bus B1, and is written in the compressed image area 40d (see FIG. 3) of the SDRAM 40. The I / F circuit 48 reads the compressed image data stored in the compressed image area 40d through the memory control circuit 38, and records the read compressed image data on the recording medium 50 in a file format. Note that the recording medium 50 is detachable, and can be accessed by the CPU 26 when mounted in a slot (not shown).

  As described above, the CCD imager 14 has the imaging surface 14f in which a plurality of pixels are two-dimensionally arranged. Raw image data representing the scene captured by the imaging surface 14 f is created by the CDS / AGC / AD circuit 16. The luminance values Laound [0] to Round [7] of the surrounding 8 pixels present around the defective pixel among the plurality of pixels are detected by the CPU 26 from the raw image data created by the CDS / AGC / AD circuit 16. The CPU 26 determines whether or not a line across the defective pixel is drawn on the partial image area represented by the defective pixel and the surrounding eight pixels based on the detected plurality of luminance values, and the luminance of the defective pixel The value DEFECT is corrected in a different manner depending on the determination result.

  The correction mode of the luminance value DEFECT of the defective pixel differs depending on whether or not a line straddling the defective pixel is drawn. Thus, accurate defective pixel correction is realized.

  When the CPU 26 is set to the photographing mode by a key input device (not shown), specifically, the CPU 26 performs processing according to the flowcharts shown in FIGS. Note that the program corresponding to this flowchart is stored in the flash memory 52.

  Referring to FIG. 11, defective pixel information is read from flash memory 52 in step S1. The read out defective pixel information is stored in the memory 26m. In step S3, through image processing is executed. As a result, a through image representing the object scene is reproduced on the LCD 46.

  In step S5, it is determined whether or not the shutter button 54 has been half-pressed. If the determination result is negative, through image AE processing is executed in step S7. As a result, the brightness of the through image displayed on the LCD 46 is appropriately adjusted. If the determination result is affirmative, a recording AE process is executed in step S9. As a result, the adjustment is made more strictly based on the luminance evaluation value fetched from the luminance evaluation circuit 24.

  In step S11, it is determined whether or not the shutter button 54 has been fully pressed. If the determination result is negative, it is determined in step S13 whether or not the shutter button 54 has been released. When the shutter button 52 is released, the process returns to step S3, and when the shutter button 54 is held, the process returns to step S11.

  When the shutter button 54 is fully pressed, shooting processing is executed in step S15. When the raw image data is stored in the SDRAM 40, pixel correction processing is executed in step S17. When the defective pixel correction process is completed, a recording process is executed in step S19. When this process is completed, the process returns to step S3. As a result, compressed image data representing the object scene image at the time when the shutter button 54 is fully pressed is recorded in the recording medium 50 in a file format.

  The pixel correction process in step S17 is executed according to a subroutine shown in FIG. In step S21, based on the position information Adr of the defective pixel registered in the memory 26m, the luminance value DEFECT of the defective pixel and the luminance values Round [0] to Round [7] of the surrounding eight pixels are detected from the raw image data. In step S23, an alignment process is executed. As a result, the aligned luminance values Lsort [0] to Lost [7] are stored in the work area 40c.

  In step S25, it is determined whether or not the partial image represented by the defective pixel and the surrounding eight pixels is a low-contrast image. If the partial image represented by the defective pixel and the surrounding 8 pixels is a low-contrast image, the process proceeds to step S33. On the other hand, if the partial image represented by the defective pixel and the surrounding 8 pixels is a high-contrast image, the process proceeds to step S27. It is determined whether or not 8 pixels have defective pixels. If the surrounding 8 pixels have defective pixels, the luminance values Laound [0] to Laound [7] of the surrounding 8 pixels are regarded as samples with low reliability for the average value filter processing, and the process proceeds to step S33. If the surrounding 8 pixels do not have a defective pixel, the brightness values around [0] to [7] of the surrounding 8 pixels are regarded as samples having high reliability with respect to the average value filtering, and the average value is obtained in step S29. Perform filtering.

  In step S31, it is determined whether or not a flag Flag described later is “0”. If the flag Flag is “1”, the process proceeds to step S35. If the flag Flag is “0”, the median filter process is executed in step S33. As a result, the luminance value DEFECT of the defective pixel is replaced with the median value described above.

  In step S35, it is determined whether or not all defective pixels have been corrected. If the determination result is negative, the process returns to step S21. If the determination result is affirmative, the process returns to the upper hierarchy routine.

  The average value filtering process in step S29 is executed according to a subroutine shown in FIG. In step S41, the flag Flag is set to “0”, the threshold value TH is calculated in step S43, and the classification process is executed in step S45. As a result, each of the aligned luminance values Lsort [0] to Lsort [7] is classified into one of the high luminance pixel and the low luminance pixel based on the threshold value TH.

  In step S47, it is determined whether or not the aligned luminance value classified into the low luminance pixels is either “2” or “3”. If a determination result is YES, it will be discriminate | determined by step S51 whether a defective pixel and surrounding 8 pixels correspond to the exclusion pattern 1 (refer FIG. 5 (A)-(P)). If the defective pixel and the surrounding 8 pixels correspond to the exclusion pattern 1, it is determined that a low-brightness line straddling the defective pixel is not drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. On the other hand, if the defective pixel and the surrounding 8 pixels do not correspond to the exclusion pattern 1 while returning to the routine, a low-brightness line across the defective pixel is drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. In step S55, the average value of the aligned luminance values classified into the low luminance pixels is calculated, and the luminance value DEFECT of the defective pixel is replaced with the calculated average value. As a result, the defective pixel has a luminance value that approximates the low luminance pixel formed by the line represented by the low luminance value.

  On the other hand, if the determination result in the step S47 is NO, it is determined whether or not the aligned luminance value classified into the high luminance pixel in the step S49 is either “2” or “3”. If the determination result is negative, it is determined that the aligned luminance value classified as the low luminance pixel is any one of “1”, “4”, and “7”, and the process returns to the upper layer routine. .

  If the determination result is affirmative, it is determined in step S53 whether or not the defective pixel and the surrounding eight pixels correspond to the exclusion pattern 2 (see FIGS. 8A to 8P). If the defective pixel and the surrounding 8 pixels correspond to the exclusion pattern 2, it is determined that a high-brightness line straddling the defective pixel is not drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. On the other hand, if the defective pixel and the surrounding 8 pixels do not correspond to the exclusion pattern 2 while returning to the routine, a high-brightness line across the defective pixel is drawn on the partial image represented by the defective pixel and the surrounding 8 pixels. In step S57, the average value of the aligned luminance values classified as high luminance pixels is calculated, and the luminance value DEFECT of the defective pixel is replaced with the calculated average value. As a result, the defective pixel has a luminance value that approximates the high luminance pixel formed by the line represented by the high luminance value.

  In step S59, the flag Flag is set to “1”, and the process returns to the upper layer routine. As a result, median filter processing is prohibited and defective pixels are corrected quickly.

It is a block diagram which shows the structure of one Example of this invention. It is an illustration figure which shows an example of the imaging surface 14f formed in the CCD imager 14 applied to the FIG. 1 Example. It is an illustration figure which shows an example of the memory mapping of SDRAM40 applied to FIG. 1 Example. It is an illustration figure which shows another example of the imaging surface 14f formed in the CCD imager 14 applied to the FIG. 1 Example. (A) is an illustrative view showing an example of exclusion pattern 1, (B) is an illustrative view showing another example of exclusion pattern 1, and (C) is an illustrative view showing another example of exclusion pattern 1. (D) is an illustrative view showing still another example of the exclusion pattern 1, (E) is an illustrative view showing another example of the exclusion pattern 1, and (F) is another example of the exclusion pattern 1. (G) is an illustrative view showing yet another example of the exclusion pattern 1, (H) is an illustrative view showing another example of the exclusion pattern 1, and (I) is an exclusion pattern. (J) is an illustrative view showing yet another example of the exclusion pattern 1, (K) is an illustrative view showing another example of the exclusion pattern 1, L) is an illustrative view showing another example of exclusion pattern 1. (M) is an illustrative view showing still another example of the exclusion pattern 1, (N) is an illustrative view showing another example of the exclusion pattern 1, and (O) is an illustrative view showing another example of the exclusion pattern 1. (P) is an illustrative view showing still another example of the exclusion pattern 1. (A) is an illustration figure which shows an example before the pixel correction process of the defective pixel which exists in a high frequency image area, (B) is an illustration figure which shows an example after the pixel correction process of the defective pixel which exists in a high frequency image area FIG. (A) is an illustration figure which shows an example before the pixel correction process of the defective pixel which exists in a low frequency image area, (D) is an example after the pixel correction process of the defective pixel which exists in a low frequency image area It is an illustration figure shown. (A) is an illustrative view showing an example of the exclusion pattern 2, (B) is an illustrative view showing another example of the exclusion pattern 2, and (C) is an illustrative view showing another example of the exclusion pattern 2. (D) is an illustrative view showing still another example of the exclusion pattern 2, (E) is an illustrative view showing another example of the exclusion pattern 2, and (F) is another example of the exclusion pattern 2. (G) is an illustrative view showing yet another example of the exclusion pattern 2, (H) is an illustrative view showing another example of the exclusion pattern 2, and (I) is an exclusion pattern. (J) is an illustrative view showing yet another example of the exclusion pattern 2, (K) is an illustrative view showing another example of the exclusion pattern 2, (J) L) is an illustrative view showing another example of the exclusion pattern 2. M is an illustrative view showing still another example of the exclusion pattern 2, (N) is an illustrative view showing another example of the exclusion pattern 2, and (O) is an illustrative view showing another example of the exclusion pattern 2. And (P) is an illustrative view showing still another example of the exclusion pattern 2. (A) is an illustration figure which shows another example before the pixel correction process of the defective pixel which exists in a high frequency image area, (B) is another figure after the pixel correction process of the defective pixel which exists in a high frequency image area It is an illustration figure which shows an example. (A) is an illustration figure which shows an example before the pixel correction process of the defective pixel which exists in a low frequency image area, (B) is an example after the pixel correction process of the defective pixel which exists in a low frequency image area It is an illustration figure shown. It is a flowchart which shows a part of operation | movement of CPU26 applied to the FIG. 1 Example. It is a flowchart which shows a part of other operation | movement of CPU26 applied to the FIG. 1 Example. It is a flowchart which shows a part of other operation | movement of CPU26 applied to the FIG. 1 Example.

Explanation of symbols

10 ... Digital camera 14 ... CCD imager 16 ... CDS / AGC / AD
50 ... Flash memory

Claims (9)

Imaging means having an imaging surface in which a plurality of pixels are two-dimensionally arranged;
Creating means for creating image data representing an object scene captured on the imaging surface;
Detecting means for detecting data values of a plurality of peripheral pixels existing around a defective pixel among the plurality of pixels from the image data created by the creating means;
Discriminating means for discriminating whether or not a line straddling the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels based on the plurality of data values detected by the detecting means. And an electronic camera comprising correction means for correcting the data value of the defective pixel in a different manner according to the determination result of the determination means. Classifying means for classifying the plurality of peripheral pixels into low luminance pixels and high luminance pixels,
The determining means is a first pattern determining means for determining whether or not the plurality of peripheral pixels are along a predetermined first pattern, and determines whether or not the plurality of peripheral pixels are along a predetermined second pattern. 2. The electronic camera according to claim 1, further comprising: an activation unit that activates one of the first pattern determination unit and the second pattern determination unit based on a classification result of the two pattern determination unit and the classification unit.   The correction unit performs a correction process with reference to only the low luminance pixel among the low luminance pixel and the high luminance pixel when the determination result of the first pattern determination unit is negative, 3. A second correction unit that executes a correction process with reference to only the high luminance pixel of the low luminance pixel and the high luminance pixel when the determination result of the second pattern determination unit is negative. The electronic camera described.   The correction means corrects with reference to both the low-luminance pixel and the high-luminance pixel when either one of the discrimination result of the first pattern discrimination means and the discrimination result of the second pattern discrimination means is affirmative. The electronic camera according to claim 3, further comprising third correction means for executing processing. Holding means for holding pixel information for identifying the defective pixel;
The electronic camera according to claim 1, wherein the detection unit executes a detection process based on pixel information held by the holding unit.   6. The electronic camera according to claim 5, wherein the pixel information of the defective pixel held by the holding unit includes position information indicating a position where the defective pixel appears and brightness information indicating the brightness of the defective pixel.   A color filter having a plurality of color elements covering each of the plurality of pixels; each of the plurality of color elements having one of a plurality of colors; The electronic camera according to claim 1, wherein the electronic camera matches a color of a color filter covering a defective pixel. To a processor of an electronic camera comprising an imaging means having an imaging surface in which a plurality of pixels are arranged two-dimensionally, and a creation means for creating image data representing an object scene captured on the imaging surface,
A detecting step of detecting data values of a plurality of peripheral pixels existing around a defective pixel among the plurality of pixels from the image data created by the creating unit;
A determination step of determining whether a line straddling the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels based on the plurality of data values detected by the detection step. And an image processing program for executing a correction step of correcting the data value of the defective pixel in a different manner according to the determination result of the determination step. An image processing method for an electronic camera comprising: an imaging unit having an imaging surface in which a plurality of pixels are two-dimensionally arranged; and a creation unit that creates image data representing an object scene captured on the imaging surface,
A detecting step of detecting data values of a plurality of peripheral pixels existing around a defective pixel among the plurality of pixels from the image data created by the creating unit;
A determination step of determining whether a line straddling the defective pixel is drawn on the partial image represented by the defective pixel and the plurality of peripheral pixels based on the plurality of data values detected by the detection step. An image processing method comprising: a correction step of correcting the data value of the defective pixel in a different manner depending on the determination result of the determination step.

Images