JP2013012983A - Imaging apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simply remove horizontal line noise caused by power fluctuation etc.SOLUTION: An imaging apparatus comprises: an imaging part which outputs image data photoelectrically converted by plural pixels arranged in matrix; a random number generation part which generates random numbers; a reading part which in accordance with the random numbers generated by the random number generation part, designates a pixel position on the matrix in the imaging part, and reads the image data therefrom; and an address addition part which adds, to image data to be read by the reading part, address data indicating the pixel position where the image data is read out.

Description

  The present invention relates to an imaging apparatus.

  In recent years, an image sensor used in an electronic camera or the like has a problem that streak noise, shading, or the like occurs due to various factors. Therefore, in the case of fixed noise or shading, a process of subtracting separately acquired correction data from image data for each shooting is performed. In general, such correction data is acquired in a light-shielded state, but a technique for acquiring correction data equivalent to the light-shielded state in order to reduce an error due to a characteristic difference or the like has been studied (for example, Patent Documents). 1).

JP 2008-148082 A

  However, the conventional method has a problem in that a correction process using a memory or correction data that holds the acquired correction data is required. Further, the horizontal stripe noise due to power supply fluctuation used for AD conversion of the image sensor cannot be removed by the process of acquiring correction data and subtracting it from the image data as in the prior art because the position differs for each photographing. As a countermeasure against such power supply fluctuation, use of a capacitor having a large capacity, a regulator, and a filter can be considered, but there is a problem that the number of external parts increases.

  An object of the present invention is to provide an imaging apparatus that can disperse noise in a captured image and make it difficult to see by devising a method for reading from an imaging element.

  An imaging apparatus according to the present invention includes an imaging unit that outputs image data photoelectrically converted by a plurality of pixels arranged in a matrix, a random number generation unit that generates a random number, and a random number generated by the random number generation unit A reading unit for specifying the pixel position on the matrix of the imaging unit and reading the image data, and adding address data indicating the pixel position from which the image data has been read to the image data read by the reading unit And an address adding unit.

  The image pickup apparatus according to the present invention can make it difficult to see the noise of a captured image by devising a reading method from the image pickup element.

It is a figure which shows the structural example of the electronic camera 101 which concerns on this embodiment. 2 is a diagram illustrating a configuration example of an image sensor 104. FIG. It is a figure which shows the example of an output data format. It is a flowchart which shows an imaging | photography process. It is a figure which shows the example of a timing of normal reading and random reading. It is a figure which shows the mode of horizontal stripe noise in the case of selecting a row at random. It is a figure which shows the example which selects a block at random. It is a figure which shows the mode of horizontal stripe noise in the case of selecting a block at random. It is a figure which shows the example which selects a block at random from a continuous-shot image. It is a figure which shows pixel arrangement | positioning of a color image. It is a figure which shows the example of the output data format of a color image. It is a figure which shows a simulation result.

  Hereinafter, embodiments of an imaging apparatus according to the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram illustrating a configuration example of an electronic camera 101 according to the present embodiment. In FIG. 1, an electronic camera 101 includes an optical system 102, a mechanical shutter 103, an image sensor 104, an image buffer 105, an image processing unit 106, a control unit 107, a memory 108, a display unit 109, and an operation. Section 110 and a memory card IF (interface) 111.

  In FIG. 1, the subject light incident on the optical system 102 is imaged on the light receiving surface of the image sensor 104 via the mechanical shutter 103. Here, the optical system 102 includes a plurality of lenses such as a zoom lens and a focus lens, a lens driving unit, an aperture mechanism, and the like. The control unit 107 controls the zoom lens position, the focus lens position, the aperture value, and the like. The

  The image sensor 104 is configured by, for example, a CMOS solid-state image sensor, and converts it into an electrical signal corresponding to the amount of received light by a photoelectric conversion unit of pixels arranged in a matrix on the light receiving surface. The analog electrical signal output for each pixel is A / D converted and output as digital image data. Note that, in the electronic camera 101 according to the present embodiment, as will be described in detail later, the image sensor 104 is provided with a function for making it difficult to see horizontal stripe noise due to power supply fluctuation or the like.

  The image buffer 105 is composed of, for example, a volatile high speed memory. The image buffer 105 not only temporarily stores image data output from the image sensor 104, but is also used as a buffer memory when the image processing unit 106 performs image processing. Alternatively, an image at the time of shooting or a shot image stored in the memory card 111 a connected to the memory card IF 111 is read into the image buffer 105 to perform image processing, or when an image is displayed on the display unit 109. Also used as display memory.

  The image processing unit 106 performs image processing such as white balance processing, color interpolation processing, gamma correction processing, saturation enhancement processing, and contour outline enhancement processing on the image data captured in the image buffer 105. Further, the image processing unit 106 performs compression processing on the image data after image processing using an image compression method compliant with the JPEG standard or the like. Furthermore, as will be described later, the electronic camera 101 according to the present embodiment performs image composition processing for compositing a captured image based on address data added to image data read from the image sensor 104.

  The control unit 107 is configured by a CPU that operates according to a program stored therein, for example. The control unit 107 controls the entire operation of the electronic camera 101. For example, the shooting mode selection dial 111a of the operation unit 110 selects the shooting mode (continuous shooting, single shooting, etc.) of the electronic camera 101, or the release button 111b. When the button is pressed, lens control and aperture control of the optical system 102 are performed to open and close the mechanical shutter 103, and a subject image is captured by the image sensor 104. Then, the control unit 107 performs control to read out the image data from the image sensor 104 and load it into the image buffer 105. Further, the control unit 107 instructs the image processing unit 106 to perform predetermined image processing on the image data captured in the image buffer 105, and adds a predetermined file name to the image data after image processing (for example, JPEG data). And header information is added and stored in the memory card 111a. Alternatively, the image data captured in the image buffer 105 is displayed on the display unit 109 as a captured image.

  The memory 108 is configured by a non-volatile semiconductor memory such as a flash memory, for example. Parameters such as a shooting mode, exposure information, and focus information of the electronic camera 101 are stored. The control unit 107 controls the operation of the electronic camera 101 with reference to these parameters.

  The display unit 109 is composed of a liquid crystal monitor or the like. Then, the control unit 107 displays a captured image, a setting menu necessary for operating the electronic camera 101, and the like.

  The operation unit 110 includes a shooting mode selection dial 111a, a release button 111b, and the like. The user operates these operation buttons to use the electronic camera 101. For example, the shooting mode selection dial 111a selects continuous shooting or single shooting. Note that the operation information from these operation units 110 is output to the control unit 107, and the control unit 107 controls the operation of the electronic camera 101 in accordance with the operation information input from the operation unit 110.

  The memory card IF 111 is an interface for connecting the memory card 111 a to the electronic camera 101. The control unit 107 reads and writes image data to and from the memory card 111a via the memory card IF 111.

The above is the configuration and basic operation of the electronic camera 101 according to the present embodiment.
[Configuration of Image Sensor 104]
Next, the configuration of the image sensor 104 for realizing a function for making it difficult to see the horizontal stripe noise due to power supply fluctuation, which is a feature of the electronic camera 101 according to the present embodiment, will be described in detail.

  FIG. 2 is a diagram illustrating a configuration example of the image sensor 104. In FIG. 2, the image sensor 104 includes a pixel unit 201, a scanning circuit 202, a readout circuit 203, an A / D conversion unit 204, a readout control unit 205, a random number generation unit 206, and an address addition unit 207. Have.

  The pixel unit 201 includes a plurality of pixels P arranged in a matrix, and reads out an electric signal photoelectrically converted by the pixel P in accordance with a timing signal output from the scanning circuit 202 to the reading circuit 203. In FIG. 2, the pixel unit 201 has pixels P in N rows and M columns (N and M are natural numbers). For example, a pixel in the first row and the second column is described as a pixel P (1, 2). Here, when x is an integer from 1 to N and y is an integer from 1 to M, (x, y) of the pixel P (x, y) indicates pixel coordinates. Here, in the electronic camera 101 according to the present embodiment, when image data is read from the image sensor 104, a row or a pixel P is selected at random, and the horizontal stripe noise can be made invisible. In the present embodiment, the read control unit 205, the random number generation unit 206, and the address addition unit 207 are provided inside the image sensor 104, but may be provided outside the image sensor 104.

  The scanning circuit 202 selects a pixel (or row or region) from which a signal is read from the pixel unit 201 and outputs a timing signal for reading a signal from the selected pixel P. For example, the first row is selected, and a signal is read from each pixel (pixel P (1,1 to pixel P (1, M)) of the first row to the reading circuit 203. The scanning circuit 202 A timing signal for outputting the signals for one row read out and stored in 203 in pixel units in the column order is output.

  The readout circuit 203 includes a circuit that temporarily holds a signal read from each pixel P by the scanning circuit 202. The signal held in the readout circuit 203 is output to the A / D conversion unit 204 in units of pixels in accordance with the timing signal output from the scanning circuit 202.

  The A / D converter 204 converts an analog electric signal output from the readout circuit 203 into digital image data.

  The readout control unit 205 reads out an electrical signal from the pixel unit 201 to the scanning circuit 202 based on a clock signal, a vertical synchronization signal (VD), a horizontal synchronization signal (HD), or a readout start signal output from the control unit 107. Indicates the position. At this time, the read control unit 205 selects a pixel position to be read according to the random number input from the random number generation unit 206. Further, the read control unit 205 outputs address data indicating the pixel position of the image data read by the scanning circuit 202 from the pixel unit 201 and the read circuit 203 to the address adding unit 207.

  The random number generation unit 206 generates a random number according to a command from the read control unit 205 and outputs the random number to the read control unit 205. As a random number generation method, a generator polynomial by an M-sequence code may be configured by hardware such as a flip-flop circuit to generate a pseudo-random number, or by software such as a uniform random number generation method such as the Mersenne Twister method. A random number may be generated. Note that the generated random number corresponds to the number of rows and pixel coordinates constituting the pixel unit 201 or the number of blocks in the case of a block method described later. In addition, when the generated random number values are duplicated, they are removed, and the processing is continued until all values corresponding to the number of rows are generated. For example, in the pixel unit 201 of FIG. 2, when a row is selected at random, the random number generation process is continued until all the values from 1 to N have been generated.

  The address addition unit 207 adds the address data output by the read control unit 205 to the image data converted by the A / D conversion unit 204 and outputs the image data from the image sensor 104. Here, a method for adding address data to image data will be described. FIG. 3A is a diagram illustrating a format example of data output from a general imaging device. In the case of FIG. 3A, the image data 502 of each pixel is output between the start code 501 and the end code 503 for synchronization. Since a normal image sensor reads out sequentially from the first line, a captured image can be reproduced without adding information indicating which line the image data 502 is.

  On the other hand, since the electronic camera 101 according to the present embodiment randomly selects rows and pixels P in order to make the horizontal stripe noise less visible as described above, the electronic camera 101 is read from the image sensor 104. In order to reproduce a captured image from image data, information (address data) indicating which position (row, pixel, or block position) the image data 502 is in the pixel unit 201 is required. Therefore, the address adding unit 207 outputs address data 504 and image data 502b between the start code 501 and the end code 503 as shown in FIG. For example, FIG. 3C shows an example of output data when row position data 505 is added as address data 504 when image data is read out by randomly selecting a row of the pixel unit 201. As a result, it is possible to determine whether the read image data is the image data of the first row or the image data of the Nth row, and the control unit 107 and the image processing unit 106 of the electronic camera 101 can detect the image buffer 105. The captured image data having the same pixel arrangement as that of the pixel portion 201 of the image sensor 104 can be generated by rearranging the image data of each row captured in the image sensor 104.

[Processing during shooting]
Next, processing at the time of shooting in the electronic camera 101 according to the present embodiment will be described using the flowchart of FIG. 4 is a process executed according to the program of the control unit 107, and is a process when the release button 110b of the operation unit 110 is pressed. When the control unit 107 detects that the release button 110b is pressed, the control unit 107 outputs an image data read start signal to the read control unit 205 of the image sensor 104.

  (Step S101) The read controller 205 causes the random number generator 206 to generate a random number.

  (Step S102) The read control unit 205 selects the read position (row position, pixel position, block position, etc.) of the pixel unit 201 corresponding to the random number generated by the random number generation unit 206.

  (Step S <b> 103) The scanning circuit 202 reads a signal from the pixel at the reading position selected by the reading control unit 205 to the reading circuit 203, and further converts the signal into image data by the A / D conversion unit 204.

  (Step S <b> 104) The address adding unit 207 adds the address data output by the read control unit 205 corresponding to the image data converted by the A / D conversion unit 204 and outputs the address data to the image buffer 105.

  (Step S <b> 105) The read control unit 205 determines whether signals have been read from all rows or pixels of the pixel unit 201. If all reading has been completed, the process proceeds to step S106. If all reading has not been completed, the process returns to step S101 to repeat the same processing.

  (Step S106) The control unit 107 instructs the image processing unit 106 to rearrange the image data with address data fetched into the image buffer 105 at the position indicated by the address data, and is the same as the pixel unit 201 of the image sensor 104. One shot image with pixel arrangement is synthesized. Note that the synthesized image is held in the image buffer 105.

  (Step S <b> 107) The control unit 107 stores the image data combined by the image processing unit 106 held in the image buffer 105 in the memory card 111 a connected via the memory card IF 111. Actually, an ordinary electronic camera such as white balance processing, color interpolation processing, gamma correction processing, saturation emphasis processing, contour contour emphasis processing, or image compression processing such as JPEG is applied to the synthesized image data. The same process is performed, a predetermined file name determined in advance is added, and the file is stored in the memory card 111a.

  Here, the image data with address data for one screen held in the image buffer 105 may be stored in the memory card 111a as one file without performing the combining process in step S106. In this case, application software that performs the same composition processing as in step S105 is installed in the personal computer, and after shooting with the electronic camera 101, the memory card 111a is connected to the personal computer and the application software that performs the composition processing is launched. To perform image composition processing. As a result, the electronic camera 101 does not need to perform a composition process at the time of shooting, and the processing load can be reduced.

[Example when image data is read out by randomly selecting a row]
Next, an example of selecting and reading out rows from the pixel unit 201 of the image sensor 104 at random will be described. In a normal imaging device, as shown in FIG. 5A, image data is read in order from the first row of the pixel unit 201. Here, VD (vertical synchronization signal) is a synchronization signal output for each screen, and HD (horizontal synchronization signal) is a timing signal output for each row. In the case of FIG. 5 (a), up to the last 3000th row is read out in order of the first row, the second row, the third row,... In synchronization with the HD. In the example of FIG. 5A, it is assumed that the pixel unit 201 has 3000 rows. On the other hand, in the image sensor 104 of the electronic camera 101 according to the present embodiment, as shown in FIG. 5B, the row corresponding to the random number generated by the random number generator 206 is selected at random, and the image data is selected. read out. Here, VD and HD are the same as those in FIG. For example, in the case of FIG. 5B, image data is read from all rows in the order of rows selected at random, such as the fourth row, the 1010th row, the 501st row,. In the case of FIG. 5B as well, it is assumed that the pixel portion 201 has 3000 rows as in FIG. The random number generator 206 continues to generate random numbers until line numbers corresponding to all 3000 lines are generated. If the generated random numbers correspond to overlapping line numbers, the random number is generated and the next random number is generated. To do. Note that the random number generation unit 206 may simply generate random numbers and ignore the overlapping random numbers on the read control unit 205 side. In the example of FIG. 5B, reading is started from the 4th row, the 1010th row, the 501st row,..., The 901st row, the 1501st row, the 701st row, and the 10th row for all 3000 rows. Reading of the image data from is terminated. Note that the random number generation unit 206 performs random number generation processing at a higher speed than the HD timing, so even if the processing is performed again to generate random numbers with duplicate line numbers, the random number generation unit 206 does not deviate from the HD timing.

  Here, the horizontal stripe noise will be described. When the image data is read out in the row order as shown in FIG. 5A, the horizontal stripe noise 301a as shown in the image 301 of FIG. 6 appears in the image due to the power supply fluctuation of the A / D converter 204 or the like. Normally, the horizontal streak noise 301a appears conspicuously because it appears in a plurality of lines as drawn in a dotted circle 351. For example, in the case of FIG. 5A, the horizontal stripe noise is placed in the first to third lines and the 2997th to 2999th lines in which the noise indicated by the dotted ellipse over a plurality of lines is generated.

  On the other hand, since the electronic camera 101 according to the present embodiment reads image data from a randomly selected row as described with reference to FIG. 5B, noise is generated as in FIG. Even in this case, the noise is distributed to the fourth line, the 1010th line, the 501st line, the 901st line, the 1501st line, and the 701st line. Thereby, as shown in the image 302 of FIG. 6, the horizontal streak noise 302 a becomes only one line as depicted in the dotted circle 352, and becomes less conspicuous. This corresponds to, for example, one of 3000 lines. For example, when the length in the vertical direction when displayed on the display unit 109 or printed on photographic paper is 90 mm, the width of one line is 0.03 mm. Not too much. Furthermore, in JPEG image compression processing or the like, since it is blocked and irreversible compression processing is performed, it becomes even less conspicuous. In the example of FIG. 6, the horizontal stripe noise 301a and the horizontal stripe noise 302a are exaggerated so that the effect of the electronic camera 101 according to the present embodiment can be easily understood.

[Example of selecting block randomly and reading image data]
Next, an example in which the captured image is divided into a plurality of blocks, the block positions are selected at random, and image data is read out for each block will be described. In this case as well, the basic configuration of the electronic camera 101 and the image sensor 104 is the same as in FIGS. 1 and 2, but the operation of each part is slightly different.

  FIG. 7 illustrates an example in which the captured image 401 is divided into nine blocks and read. In FIG. 7, the number of blocks is nine for easy understanding, but the number of blocks may be further divided. In the example of FIG. 7, the random number generation unit 206 generates random numbers in the order of 3, 9, 5, 7, 2, 8, 6, 1, and 4. Here, the random number generation unit 206 repeats the operation of generating the next random number by ignoring when a duplicated random number or a number not assigned to the block number (for example, 0) is generated, corresponding to all blocks. It is assumed that nine random numbers from 1 to 9 are generated.

  Then, the read control unit 205 selects the block position of the pixel unit 201 corresponding to the random number generated by the random number generation unit 206. The scanning circuit 202 reads a signal from each pixel of the block selected by the reading control unit 205 to the reading circuit 203, and further converts the signal into image data by the A / D conversion unit 204. In the example of FIG. 7, address data (3) is added to the image data read from the block (3) first. Similarly, the second block (9), the third block (5), the fourth block (7), and finally the block (4) are addressed to the respective image data read from each block. Data (9), (5), (7), (4) are added respectively.

  Here, the scanning circuit 202 reads image data from each pixel in the order of rows in the block. However, as in the previous embodiment, the order of rows in the block may be selected at random. As a result, the horizontal stripe noise is further dispersed and can be made even less conspicuous.

  The address adding unit 207 adds data (block number or the like) indicating the block position selected by the read control unit 205 to the image data in the block converted by the A / D conversion unit 204 as address data, and the image buffer 105. Output to. FIG. 3D shows an output when the block position data 506 is added to the image data 502 c as the address data 504 in FIG. 3B when the image data is read out by randomly selecting the block position of the pixel unit 201. An example of data is shown. Thereby, it is possible to determine which block of the block (1) to the block (9) the read image data 502c is, and the control unit 107 or the image processing unit 106 takes in the image buffer 105. The captured image data having the same pixel arrangement as that of the pixel portion 201 of the image sensor 104 can be generated by rearranging the image data of each block. As described above, when reading out the block in units of rows, image data in the format of FIG. 3D is output by the number of rows in the block. Here, when outputting in the order of rows, the control unit 107 and the image processing unit 106 can generate captured images by rearranging the image data of each row in one block captured in the image buffer 105 in order. When a row in the block is selected at random, it is impossible to determine which row in the block is the image data. Therefore, the read control unit 205 outputs the block position data 506 and the row position data 505a of the read image data 502c to the address adding unit 207, and the address adding unit 207 displays the image as shown in FIG. Block position data 506 and row position data 505a are added to the data 502c. Thereby, the control unit 107 and the image processing unit 106 can rearrange the image data of each row in the block captured in the image buffer 105 to an appropriate position.

  Note that the reading control unit 205 repeatedly executes the same processing until reading of image data from all the blocks of the pixel unit 201 is completed. Then, the control unit 107 instructs the image processing unit 106 to rearrange the image data with the block position data fetched into the image buffer 105 into the position indicated by the block position data, and the image and one sheet in each block. Composite the shot images.

  Here, the effect when the image data is read out by randomly selecting a block will be described with reference to FIG. In FIG. 8, the same reference numerals as those in FIG. As described above with reference to the image 301 in FIG. 6, horizontal streak noise 301 a composed of a plurality of lines appears in the image as indicated by a dotted circle 351. On the other hand, when the image data is read out by selecting a block at random, the length of the horizontal stripe noise is shortened and dispersed as in the image 303 in FIG. In the case of the image 303, the number of blocks is nine. However, as the number of blocks is increased, the length of the horizontal stripe noise becomes shorter and the horizontal stripe noise becomes inconspicuous.

  Further, the example of the image 303 shows a state in which the image data is read out in order from the inside of the block. However, as described above, when the image data is read out by randomly selecting a row in the block. As shown in the image 304, the horizontal streak noise becomes narrow as in the image 302 of FIG. Further, by dividing the block size finely, as shown in an image 305, it is possible to disperse the block to the same extent as random noise. 8 are exaggerated so that the explanation is easy to understand, but since the size of the pixels is actually very small, it can hardly be recognized that it is horizontal stripe noise.

  In this way, by dividing the photographed image into a plurality of blocks and randomly selecting a block from which the image data is read, the horizontal stripe noise appearing in the photographed image can be made inconspicuous.

[Example of continuous shooting]
Next, an example in which the above-described block division method is applied to a plurality of images taken by continuous shooting will be described. In this case, the basic configuration of the electronic camera 101 and the image sensor 104 is the same as in FIGS. 1 and 2, but the operation of each part is slightly different.

  In the example of FIG. 7 described above, one shot image is divided into a plurality of blocks, and the block position is selected at random, and image data is read out for each block. Image data is read out by randomly selecting blocks at different positions from each of the images. In this embodiment, one block is selected from each image, but a plurality of blocks may be selected.

  In the example of FIG. 9, nine images from the image 401a (first image) to the image 401i (9th image) are continuously shot according to the flow of time t. In FIG. 9, the number of blocks is nine, but it may be divided into a larger number of blocks. 9, the random number generation unit 206 generates random numbers in the order of 3, 9, 5, 7, 2, 8, 6, 1, 4 as in the case of FIG. Here, the random number generation unit 206 repeats the operation of generating the next random number by ignoring when a duplicated random number or a number not assigned to the block number (for example, 0) is generated, corresponding to all blocks. It is assumed that nine random numbers from 1 to 9 are generated.

  Then, the reading control unit 205 selects the block (3) of the pixel unit 201 corresponding to the random number generated by the random number generation unit 206 when capturing the first image 401a. The scanning circuit 202 reads a signal from each pixel of the block (3) selected by the reading control unit 205 to the reading circuit 203. Then, the image data is converted into image data by the A / D conversion unit 204, and the address data is added by the address adding unit 207 and outputted to the image buffer 105. The addition of address data is the same as in the example of FIG. Next, the reading control unit 205 selects the block (9) of the pixel unit 201 corresponding to the random number generated by the random number generation unit 206 when capturing the second image 401b. Then, the scanning circuit 202 reads a signal from each pixel of the block (9) selected by the reading control unit 205 to the reading circuit 203. Further, the image data is converted by the A / D conversion unit 204, the address data is added by the address adding unit 207, and the image data is output to the image buffer 105. Similarly, block (5) is selected when the third image 401c is taken, and block (7) is selected when the fourth image 401d is taken and read from each block of each image. Image data to which address data indicating the block position is added is fetched into the image buffer 105. Similarly, image data at different block positions are read out from the fifth image 401e to the ninth image 401i.

  In this manner, nine image data at different block positions read out to the image buffer 105 are combined into one photographed image as in the example of FIG. In this case as well, the effect described with reference to FIG. 8 can be obtained, but since the captured image itself is different, the horizontal stripe noise is further dispersed and becomes inconspicuous. Note that this example is suitable for shooting a subject with little movement during continuous shooting.

[Example of color image]
In each of the above examples, a color image is not particularly mentioned, but here, an example in which the image sensor 104 is a Bayer array color image will be described. FIG. 10A is a diagram illustrating a pixel array of an image 601 captured by the image sensor 104 having a general Bayer array. In the image 601, R pixels and Gr pixels are arranged alternately in odd rows, B pixels and Gb pixels are arranged alternately in the horizontal direction, and each row is arranged alternately in the vertical direction. Here, the R pixel is a pixel for acquiring red component image data, the B pixel is a pixel for acquiring blue component image data, and the Gr and Gb pixels are pixels for acquiring green component image data. Note that r in Gr indicates G pixel in the row of R pixels, and b in Gb indicates G pixel in the row of B pixels.

  Address data is added by the address adding unit 207 to the image data read from the pixel unit 201 of the image sensor 104 in the Bayer array to the reading circuit 203 and A / D converted by the A / D conversion unit 204. FIG. 11A shows an example of a data format in which address data is added to image data by the address adding unit 207 when a color image is read out by randomly selecting a row. In the example of FIG. 11A, the random number generation unit 206 generates random numbers in the order of 1, 8, 4, 7,..., And reads image data from the row corresponding to each random value. For example, in the case of FIG. 11A, the row position data 601 indicating the first row corresponding to the random number 1 and the image data 602 of the first row are output between the start code 501 and the end code 503 for synchronization. .

  Here, since the image data 602 is an odd row, the image data in the pixel order of Gr, R, Gr, R, Gr. Next, row position data 601 indicating the eighth row corresponding to the random number 8 and image data 602 on the eighth row are output. Here, since the image data 602a is an even-numbered row, image data in the pixel order of B, Gb, B, Gb, B... Is output. Similarly, as the image data 602b in the fourth row corresponding to the next random number 4, image data in the pixel order of B, Gb, B, Gb, B... Is output, and 7 corresponding to the next random number 7 is output. As the image data 602c in the row, image data in the pixel order of Gr, R, Gr, R, Gr,.

  In this way, it is possible to determine which row of image data the read image data is, and the control unit 107 and the image processing unit 106 of the electronic camera 101 can read the image of each row captured in the image buffer 105. Data can be rearranged to generate a captured image having the same pixel arrangement as the pixel portion 201 of the image sensor 104.

  Here, the raw image data read from each pixel of the Bayer array such as the image 601 in FIG. 10A is called raw data, and each pixel has only data of any color component of RGB. do not have. Therefore, color interpolation processing is performed so that RGB color component data can be obtained for each pixel. In general color interpolation processing, interpolation processing is performed using image data of surrounding pixels, and as shown in an image 602 in FIG. 10B, RGB color component data is generated for each pixel. Therefore, in the case of a color image with a Bayer array, for example, even if there is horizontal stripe noise in the third row of the image 601 shown in FIG. 10A, the adjacent row (in this case, the second row or the first row). ) Has no horizontal stripe noise, the B component and the Gb component of the R pixel position in the third row (for example, the R pixel position of the pixel P (3,4)) generated by the color interpolation process include the horizontal stripe noise. Therefore, when viewed as a color image, the influence of the horizontal stripe noise in the pixel P (3,4) is reduced as compared with the case of a monochrome image.

[Modification of color image]
In FIG. 11A, since the random readout is performed for each row, pixels of different color components in the same row also contain the same horizontal stripe noise. However, in this modification, a row is selected at random, Read in multiple times for each color.

  FIG. 11B shows an example of a data format in which the address adding unit 207 adds address data to image data when image data is read for each color from a randomly selected row. In the example of FIG. 11B, the random number generation unit 206 generates random numbers in the order of 1, 8, 4..., And reads the image data by dividing into two times for each color from the row corresponding to each random value. In FIG. 11B, line position data 601d and image data 602d including information indicating the first line corresponding to the random number 1 and the color component (Gr) between the start code 501 and the end code 503 for synchronization. Are output. In this case, as the image data 602d, image data of Gr component pixels in an odd-numbered column of Gr, Gr, Gr. Subsequently, row position data 601e and image data 602e including information indicating the same first row and color component (R) are output. In this case, as the image data 602e, image data of R component pixels in even-numbered columns of R, R, R.

  Similarly, row position data 601f and image data 602f including information indicating the eighth row corresponding to the random number 8 and the color component (B) are output. In this case, as the image data 602f, the image data of the B component pixels in the odd-numbered columns B, B, B. Subsequently, line position data 601g and image data 602g including information indicating the same eighth line and color component (Gb) are output. In this case, as the image data 602g, image data of Gb component pixels in an even-numbered column of Gb, Gb, Gb.

  As a result, it is possible to determine which color component of which row the read image data is, and the control unit 107 and the image processing unit 106 of the electronic camera 101 can read the image of each row captured in the image buffer 105. Data can be rearranged to generate a captured image having the same pixel arrangement as the pixel portion 201 of the image sensor 104.

  As described above, in this modified example, as shown in the image 601 in FIG. 10A, even if there is horizontal stripe noise in the R pixel in the third row, the Gr pixel in the same row or the adjacent row (in this case) Is the second line or the first line) If the B and Gb pixels do not contain horizontal stripe noise, the R of the third line generated by the color interpolation processing when generating the image 602 of FIG. Since data other than the R component (B component, Gb component, and Gr component) of the pixel position (for example, the R pixel position of the pixel P (3, 4)) does not include horizontal stripe noise, the pixel P when viewed as a color image The influence of horizontal stripe noise in (3, 4) is further reduced than in the previous example.

  In the example of the color image described above, the case where the rows are selected at random is shown, but the block division method described above may be applied to the color image. In this case, the rows may be selected randomly for each block, or the image data may be read out in a plurality of times for each color component included in the randomly selected rows.

  Here, the noise reduction effect of the electronic camera 101 according to the present embodiment will be described. FIG. 12 is a graph showing the magnitude of noise when a Bayer color image captured with light shielding is divided into 18 blocks and randomly read, and the magnitude of noise when normal reading is performed as usual. This is a simulation result. FIG. 12A shows the case of normal reading, and FIG. 12B shows the case of pseudo-random reading. In FIGS. 12A and 12B, the vertical axis indicates the output value (LSB) of the pixel, and the horizontal axis indicates the vertical address (row position). The left side of each graph corresponds to the upper part of the image and the right side corresponds to the lower part of the image.

  It can be seen that the fluctuation range of the output value is smaller in the case of random reading shown in FIG. 12B than in the case of normal reading shown in FIG. That is, horizontal stripe noise in the captured image is reduced. In particular, in the case of a color image, not only the horizontal streak noise is dispersed and becomes difficult to see, but also as described above, the noise dispersed by random readout is further increased by the color interpolation processing of the Bayer array or the image compression processing such as JPEG. There is also an effect that it is reduced.

  As described above, the electronic camera 101 according to the present embodiment can make it difficult to see the horizontal stripe noise of the captured image by randomly reading the image data from the image sensor 104.

  As described above, the imaging apparatus according to the present invention has been described by way of example in each embodiment. However, the imaging apparatus can be implemented in various other forms without departing from the spirit or main features thereof. Therefore, the above-described embodiment is merely an example in all respects and should not be interpreted in a limited manner. The present invention is defined by the claims, and the present invention is not limited to the text of the specification. Further, all modifications and changes belonging to the equivalent scope of the claims are within the scope of the present invention.

DESCRIPTION OF SYMBOLS 101 ... Electronic camera; 102 ... Optical system; 103 ... Mechanical shutter; 104 ... Imaging element; 105 ... Image buffer; 106 ... Image processing part; DESCRIPTION OF SYMBOLS 108 ... Memory; 109 ... Display part; 110 ... Operation part; 110a ... Shooting mode selection dial; 111b ... Release button; 112 ... Memory card IF; 201 ... pixel part; 202 ... scanning circuit; 203 ... readout circuit; 204 ... A / D conversion part; 205 ... readout control part; 206 ... random number generation part; ..Address adding part

Claims (7)

An imaging unit that outputs image data photoelectrically converted by a plurality of pixels arranged in a matrix; and
A random number generator for generating random numbers;
A reading unit that reads out the image data by designating a pixel position on a matrix of the imaging unit according to the random number generated by the random number generation unit;
An image pickup apparatus comprising: an address adding unit that adds address data indicating the pixel position from which the image data is read to the image data read by the reading unit. The imaging device according to claim 1,
The reading unit reads the image data in units of rows by specifying the row of the imaging unit according to the random number generated by the random number generation unit,
The imaging apparatus, wherein the address adding unit adds address data indicating the row position from which the image data is read to the image data read by the reading unit. The imaging device according to claim 1,
Dividing the imaging area of the imaging unit into a plurality of preset blocks;
The reading unit reads the image data in block units by designating the block of the imaging unit according to the random number generated by the random number generation unit,
The address adding unit adds address data indicating the block position from which the image data is read to the image data read by the reading unit. The imaging device according to claim 3.
When the imaging unit continuously captures a plurality of images,
The image reading apparatus, wherein the reading unit reads the image data in units of blocks by designating the block of the image pickup unit according to the random number generated by the random number generation unit for each continuous shooting. In the imaging device according to any one of claims 1 to 4,
An image pickup apparatus, further comprising: a combining unit that combines the image data to which the address data output from the address adding unit is added so as to correspond to the position of the image pickup unit indicated by the address data. The imaging device according to claim 2,
In the imaging unit, pixels of different color components are alternately arranged in the same row,
When the image data is read from the imaging unit in units of rows, the reading unit reads out pixels in the same row for each same color component in a plurality of times. The imaging device according to claim 3.
In the imaging unit, pixels of different color components are alternately arranged in the same row,
The reading device reads the pixels in the same block for each of the same color components in a plurality of times when reading the image data from the image pickup unit in units of blocks.

Images