JP5680520B2 - Image processing apparatus and image processing method - Google Patents

Description

  Embodiments described herein relate generally to an image processing apparatus and an image processing method for reading an image such as a scanner.

  Conventionally, when reading a document with a color copier, a color scanner, a color facsimile, or a composite device (MFP) that combines them, mechanical vibration due to printing operation, impact from the outside, or paper transport in an automatic paper feeder The RGB color line sensor may not be able to read the original at the correct position due to the influence of the roller contact or vibration caused by the pulling of the roller. The positional deviation between the RGB line sensors that occurs at the time of reading becomes a cause of color deviation in reading a black line or the like.

In order to reduce this displacement, various image processing apparatuses that correct image data read by an RGB line sensor have been proposed. For example, using any one of the RGB color data as a reference color, the density difference between the reference color and the color other than the reference color at the same pixel position of the target pixel, and the position that includes virtually one pixel or less other than the reference color are shifted. An image correction unit is disclosed that calculates a density difference between a target pixel and a reference color at the same pixel position, and selects an optimum one from the calculated plurality of density differences to correct image data. (See Patent Document 1)
However, there is a problem that the positional deviation to be corrected is only in the main scanning direction with respect to the reading direction, and the positional deviation in the sub-scanning direction cannot be corrected.

JP 2005-341417 A

  The problem to be solved by the present invention is to solve the above problems and provide an image processing apparatus and an image processing method for correcting misalignment occurring at the time of reading by a line sensor in both main scanning and sub scanning directions. It is to be.

In order to achieve the above object, the image processing apparatus according to the present embodiment uses the output signal of a line sensor unit including a plurality of line sensors including at least a color line sensor R, a line sensor G, and a line sensor B as pixels. Means for acquiring in units; means for generating a reference color signal using one or more of output signals of the plurality of line sensors; and generating a processing unit array including a pixel of interest for image processing for the reference color signal Means for setting a reference color correlation array including the target pixel and having a smaller number of array elements than the processing unit array, and a scanning range in the processing unit array with the same number of array elements as the reference color correlation array Means for extracting a correlation array group of color colors included in the color code; means for calculating a correlation between the correlation array of the reference color and each correlation array included in the correlation array group of the color color; Means for calculating a correlation array of color colors having the highest correlation with the correlation array of reference colors, and pixel values included in the image correlation array of color colors having the highest correlation with the pixel values included in the correlation array of reference colors And a means for generating a color pixel value for the pixel of interest using the regression line.

1 is an overall view of an image processing apparatus according to a first embodiment. The conceptual diagram of the color sensor in the embodiment. The block diagram of the image processing part in the embodiment. 2 is a configuration example of a reference color calculation unit in the embodiment. 2 is a configuration example of a pixel array generation unit in the same embodiment. Configuration example of correlation calculation unit in same embodiment 2 is a configuration example of a color misregistration correction unit in the embodiment. 6 is a flowchart of image processing in the embodiment. Explanatory drawing of the processing unit arrangement | sequence of the reference signal and attention pixel in the embodiment. Explanatory drawing of the correlation arrangement | sequence in the same embodiment. FIG. 5 is a view for explaining extraction of a correlation array of color colors in the embodiment. Explanatory drawing of the correlation coefficient and regression line in the embodiment. (A) is when the correlation is high, and (b) is when the correlation is low. Explanatory drawing of the regression line used for position shift correction of the embodiment. Explanatory drawing of the position shift correction process in the embodiment. FIG. 10 is a block diagram of an image processing unit in the second embodiment. The block diagram of the reference color calculation part in the embodiment. Explanatory drawing of the position shift correction process including the resolution conversion in the embodiment.

  Hereinafter, embodiments will be described in detail with reference to FIGS. 1 to 17.

<First Embodiment>
In this embodiment, in an image reading apparatus such as a scanner, a positional deviation between RGB color signals simultaneously read by three RGB color line sensors is corrected.

  FIG. 1 shows an overall view of the image processing apparatus of the image processing apparatus according to the present embodiment. This embodiment is a color copier that performs color printing at a high resolution, and further performs various processing such as a function of scanning and copying a base paper, a function of communicating with an external device, a function of transmitting and receiving and printing as a facsimile machine. A composite device (MFP) or the like that can be performed is included.

  The image processing apparatus according to the present embodiment is broadly divided into a scanner unit 11, an image processing unit 12 that performs image processing on a line sensor output of a document read by the scanner unit 11, and prints on a copy sheet for output. It consists of a printer unit 13.

  The scanner unit 11 includes an automatic feeder 110 that automatically supplies a document, a document table 111 that reads the document, a light source 112 that illuminates the document, mirrors 113a, 113b, and 113c that guide reflected light of the document, and collects the reflected light. It has a lens 114 that emits light and a line sensor unit 115. As indicated by the arrows, the document pattern read from the document table 111 of the scanner unit 11 is irradiated by the light source 112, and the reflected light is incident on the line sensor unit 115 through the lenses 114 from the mirrors 113a, 113b, and 113c. The line sensor unit 115 photoelectrically converts this by a line sensor arranged for each color and outputs an image signal.

  The image processing unit 12 performs digital processing on the image signal read by the line sensor unit 115 and outputs the digital signal to the printer unit 13.

  The printer unit 13 includes a paper feed cassette 131, a developing drum 132, a transport roller 133, a paper discharge port 134, and the like. The printer unit 13 latent in the copy paper in the paper feed device 131 based on a signal output from the image processing unit 12. Image / development processing is performed, and the paper is discharged from the paper discharge port 134.

  FIG. 2 is a conceptual diagram of the line sensor unit 115 used in this embodiment. The line sensor unit 115 used in this embodiment includes a color line sensor R (red), a line sensor G (green), and a line sensor B (blue). The color line sensor R, line sensor G, and line sensor B have the same number of light receiving elements as the pixel pitch and the number of pixels.

  In addition to the color sensors R, G, and B, there may be a line sensor in which a monochrome line sensor K is added. As shown in FIG. 2, a line sensor K (black) having a higher resolution than the RGB color sensor is usually called a hybrid sensor.

  In the case of the hybrid sensor shown in FIG. 2, since the pitch of the light receiving elements of the monochrome line sensor K is half of the pixel pitch of the color line sensors R, G, B, the resolution is doubled. For example, if the color R, G, and B signals are 300 dpi, the monochrome signal has a resolution of 600 dpi.

  A block configuration of the image processing unit 12 according to the present embodiment will be described with reference to FIGS. In FIG. 3, a part of the scanner unit 11 is also shown for easy understanding of the flow of the image processing signal.

  The image processing unit 12 A / D (Analog to Digital) outputs signals output from the line sensor R, the line sensor G, and the line sensor B that read color signals in accordance with the scanning of the two-dimensional document performed by the scanner unit 11. An image memory unit 120 for storing color R, G, B signals and monochrome signals in units of pixels after performing conversion, shading correction, and inter-line correction for correcting the reading phase difference between the line sensors; A reference color calculator 121 that calculates a reference color S signal from the color R, G, and B signals, and scans the reference color S to calculate a correlation between the reference color S and the color R, G, and B signals. A pixel array generation unit 122 that generates a plurality of pixel arrays at different positions, a correlation calculation unit 123 that calculates a correlation between the plurality of pixel arrays and the reference color S, and a plurality of A correlation function determining unit and 124 determines the highest correlation pixel arrangement in the element array having a color shift correction unit 125 to correct the positional deviation generated during reading with highest correlation pixel array.

  Next, the reference color calculation unit 121 will be described with reference to FIG. Since the line sensor unit 115 of this embodiment does not include the monochrome line sensor K, the reference color calculation unit 121 uses the addition average calculation unit 41 to calculate the average of the color R, G, and B signals (equation (1)). This signal is calculated as a reference color signal S.

S = (R + G + B) / 3 (1)
As other modifications, the reference color S signal is generated by the linear combination of the color R signal, the G signal, and the B signal mixture ratio (α, β, γ) when the color G signal is used as the reference color signal. Also good. The linear combination ratio may be determined in consideration of the sensitivity characteristics of the line sensors R, G, and B, the wavelength band, and the like.

  When the line sensor unit 115 includes the monochrome line sensor K, the monochrome K signal may be used as the reference color S. This will be described in the second embodiment.

  As shown in FIG. 5, the pixel array generation unit 122 has M rows × N columns of pixels (M) including an image processing target pixel (hereinafter referred to as a target pixel) that is an image processing target for the reference color S signal. , N is an integer) and a processing unit array generation unit 51 that sets a pixel array (hereinafter referred to as a processing unit array S1) having a processing unit, and K rows that include the same target pixel and have fewer array elements than the processing unit array S1 A pixel array having the number of array elements of L columns (K and L are integers smaller than M and N, respectively) (hereinafter referred to as correlation array S2 for calculating correlation) is set, and the same array as this correlation array S2 A correlation array extraction unit 52 that extracts the correlation array group R2 of the color R, G, and B signals of the number of elements and the correlation array group G2 and the correlation array group B2 from the scanning range of the processing unit array S1 is provided. Note that the number of array elements M, N, K, and L is preferably set to an odd number because the target pixel is preferably the center pixel of the pixel array.

  Further, as shown in FIG. 6, the correlation calculation unit 123 calculates the correlation between each correlation sequence included in the correlation sequence group R2, the correlation sequence group G2, and the correlation sequence group B2, and the correlation sequence S2. An average value calculation unit 61 for calculating an average value of pixel values in each correlation array, a standard deviation calculation unit 62 for calculating a standard deviation of pixel values in each correlation array, and a correlation between the correlation array S2 of the reference color and the color color Deviation product calculation unit 63 for obtaining a deviation product of pixel values from array group R2, correlation array group G2, and correlation array group B2, and correlation array group R2 of reference color correlation array R2, correlation array group G2 And a correlation coefficient calculation unit 64 for obtaining a correlation coefficient with the correlation array B2.

  For the correlation, an average value, a standard deviation value, and a deviation product may be appropriately combined, but in the present embodiment, a correlation coefficient is used.

  When the correlation determination unit 124 determines the correlation using the correlation coefficient as described above, the correlation array having the largest correlation coefficient (value close to 1) is the color correlation array group. Each is selected from R2, correlated sequence group G2, and correlated sequence group B2. The color color correlation arrays selected at this time are defined as a correlation array R2_M, a correlation array G2_M, and a correlation array B2_M.

  In the color misregistration correction unit 125 illustrated in FIG. 7, for each of the selected correlation array R2_M, correlation array G2_M, and correlation array B2_M, a regression line calculation unit 71 that obtains a regression line with respect to the correlation array S2 of the reference color, A color signal generation unit 72 that corrects the color signal at the target pixel position using the regression line and generates a new color RN signal, color GN signal, and color BN signal is provided.

  The flow of image processing according to the present embodiment will be specifically described with reference to FIGS. First, in step ST801 in FIG. 8, the reference color calculation unit 121 reads the color R, G, and B signals from the image memory unit 120 in units of pixels, and, for example, adds the average from the read color R, G, and B signals. The reference color S signal is calculated and generated based on a predetermined rule such as.

  In step ST802, a processing unit array S1 that is a processing unit of image processing is set for the generated reference color S signal. FIG. 9 shows an example of the processing unit array S1 generated by the processing unit array generation unit 51. In this example, a processing unit array S1 of 5 rows × 5 columns including two pixels before and after the main scanning direction and the sub-scanning direction is set for the pixel of interest located at the center.

  Next, as shown in FIG. 10, in step ST803, a correlation array having a smaller number of array elements than the processing unit array S1 is set. In the example of FIG. 10, a correlation array of 3 rows × 3 columns is set.

  This correlation array is set for the reference color S and the colors R, G, and B, and a correlation is obtained between the reference color S and the colors R, G, and B. In the processing unit array S1 of the reference color S shown in FIG. 11, the correlation array with respect to the reference color S is a 3 × 3 pixel array including the pixel of interest, and is indicated by Area_S as indicated by a bold line frame in the drawing. This is defined as a correlation sequence S2.

  In Step ST804, a color pixel array group included in the scanning range of the processing unit array S1 is extracted with the same number of array elements as Area_S. In the example of FIG. 11, nine pixel arrays from Area_1 to Area_9 are extracted. For example, Area_1 is a pixel array that is shifted by two pixels before Area_S in both the main scanning direction and the sub-scanning direction.

  In step ST805, the extracted pixel array group is defined as a correlation array group R2, a correlation array group G2, and a correlation array group B2 for each of the colors R, G, and B. The correlation array group R2 for each color, Correlation between each correlation array included in G2 and B2 and correlation array S2 which is a reference for the scanning position is obtained.

  As described above, various methods can be considered for obtaining the correlation. In this embodiment, the correlation coefficient is used to obtain the correlation coefficient r of Area_S and Area_1 to Area_9 for each color.

The correlation coefficient r is defined as shown in the equation (2), and rR is added for the R color, rG is added for the G color, and rB is added for the B color.

  Further, i is an array element number in the correlation array, N is the number of array elements, and N = 9 if the correlation array is 3 rows × 3 columns. Rave, Gave, and Bave represent average values of pixel values in each correlation array. Therefore, it can be seen that the correlation coefficient r is obtained from the pixel value average value in the correlation array, the standard deviation value of the pixel values, and the pixel value deviation product between the correlation arrays.

  Here, the relationship between the correlation coefficient r and the regression line will be described with reference to FIG. Although the color R color is illustrated here, the same applies to the other G and B colors. FIG. 12A shows the case of the correlation arrangement having a high correlation with the reference color S for the color R color, and FIG. 12B shows the case where the correlation is not so high. The horizontal axis is the pixel value of the pixels included in the correlation array S2 (Area_S) of the reference color S, and the vertical axis is included in any one of the correlation array group R2 of color R (for example, Area_1, Area_2, etc.). The pixel values of the pixels are plotted for each array element at the corresponding position.

  As shown in FIG. 12A, when the correlation between Area_S and Area_1 is high, there is little variation in pixel values with respect to the regression line La, and the correlation coefficient takes a value close to 1. Further, as shown in FIG. 12B, when the correlation between Area_S and Area_2 is low, there is a large variation in the pixel values with respect to the regression line Lb, and the correlation coefficient is obtained in FIG. The value is lower than the correlation coefficient.

  In addition, when the position shift by vibration etc. has not arisen, the correlation of Area_S and Area_5 (center pixel) becomes the highest.

  Normally, the correlation coefficient takes a positive value, but the correlation coefficient may be negative when reading a document that has been subjected to special image processing such as inversion processing. Therefore, it is preferable to evaluate the correlation with the absolute value of the correlation coefficient shown in the equation (2).

  In step ST806, the correlation coefficient with Area_S is obtained for all Area1 to Area_9, and the correlation array Area_K of the color signal having the largest correlation coefficient (close to 1) is obtained (1 ≦ K ≦ N).

  For example, when the correlation coefficient in FIG. 12A is the largest, the reading position of the color R signal is shifted by 2 pixels before the reference signal S with respect to the main scanning direction and the sub scanning direction. Can be judged.

  Next, in step ST807, as shown in FIG. 13, the correlation array S2 (Area_S) of the reference color S and the correlation array of the color signal having the largest correlation coefficient are assigned to the pixels of the pixels included in this Area_K. The regression line R = ar · S + br (Lrk) is obtained from the pixel value of the corresponding pixel of the value and Area_S. Although FIG. 13 shows the color R, the regression line G = ag · S + bg and the regression line B = ab · S + bb are similarly obtained for the colors G and B.

  In step ST808, as shown in FIG. 14, new color RN, GN, and BN signals are generated by substituting the pixel value of the target pixel into the regression line obtained in step ST807. As indicated by a dotted circle in FIG. 13, in this process, the pixel of interest moves on the regression line Lrk, so that the newly generated color RN, GN, and BN signals are positioned relative to the reference color S. As a result of correcting the shift, the color shift is reduced.

  In the misregistration correction processing in steps ST801 to ST808, the processing unit array S1 is generated while scanning all the pixels of the read original, and the same processing is repeated.

  As described above, according to the present embodiment, a correlation coefficient between the correlation array of the reference colors and the correlation array of the color R, G, and B colors moved in both the main scanning and the sub scanning is obtained. The positional deviation of the target pixel is corrected based on the regression line having the largest correlation coefficient. For this reason, both main scanning and sub-scanning have the effect of correcting color misregistration at the same time. In particular, when a regression line is obtained using a reference color using a weighted average of R, G, and B, only a color shift can be effectively corrected without causing a change in color tone.

  Further, when the number of array elements in the processing unit array is made larger than in the example of the present embodiment, color misregistration correction can be performed even for pixels that are greatly displaced due to vibration.

(Second Embodiment)
In the present embodiment, a case where the line sensor of the scanner unit 11 includes a monochrome line sensor will be described.

  FIG. 15 shows a block configuration of the image processing unit of the present embodiment. A monochrome line sensor K is added to the scanner unit 11. The line sensor K has the same resolution as the color line sensors R, G, and B or a higher resolution. When the line sensor K is included, the reference color calculation unit 151 may use the monochrome signal K directly output from the line sensor K as the reference color S signal.

  When the resolution of the line sensor K is higher than the resolution of the color line sensors R, G, B, the resolution of the line sensor K is once set to the same resolution as that of the color line sensors R, G, B as shown in FIG. For example, a low resolution conversion unit 161 is provided in the reference color calculation unit 121.

  The image processing for correcting misalignment performs the same processing as in the first embodiment. However, the color misregistration correction unit 151 performs high resolution conversion processing using the high resolution signal of the line sensor K. That is, for example, when the resolution of the color line sensors R, G, and B is set to 300 dpi and the resolution of the monochrome line sensor K is set to 600 dpi, as shown in FIG. , Four pixels of the monochrome signal K are included. By substituting these four pixels into a regression line calculated for each of the colors R, G, B by the same procedure as in the first embodiment, high resolution conversion is performed on the low resolution color R, G, B signals. Can do.

  Therefore, according to the second embodiment, in addition to the effect of the first embodiment, there is an effect that color misregistration correction and high resolution conversion processing of color R, G, and B signals can be performed simultaneously.

  As described above, according to the embodiment of the present invention, the processing unit array indicating the image processing range includes peripheral pixels in the main scanning and sub-scanning directions with respect to the target pixel. Even at the same time, misalignment correction can be performed.

  Further, in the case of a hybrid sensor having a high-resolution monochrome line sensor, resolution conversion can be performed at the same time when misalignment correction is performed.

  Therefore, when reading a document with a color sensor, even if mechanical vibrations due to printing operations, external impacts, or vibrations due to roller contact or roller pulling during paper transport in an automatic paper feeder occur, color misregistration will occur. No image output is obtained.

  In the present embodiment, the reference color correlation array is fixed and the color color correlation array is moved within the processing unit array, but the reverse is also possible. That is, the same effect can be obtained even if the correlation array of color colors is fixed at the center of the processing unit array and the correlation array of the reference colors is moved to obtain the correlation. In this case, the correlation array group is set only for the reference color, which is advantageous in terms of processing speed and memory usage.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. For example, although the resolution of the monochrome line sensor of the hybrid sensor is set to twice the resolution of the color line sensor, the present invention is not limited to this. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

11 ... scanner part,
12: Image processing unit,
13 ... Printer section,
120: Image memory unit,
121 ... reference color calculation unit,
122 ... Pixel array generation unit,
123 ... correlation calculation unit,
124 ... correlation determination unit,
125. Color misregistration correction unit.

Claims (9)

Means for acquiring an output signal of a line sensor unit composed of a plurality of line sensors including at least a line sensor R, a line sensor G, and a line sensor B for each pixel;
Means for generating a reference color signal using one or more of the output signals of the plurality of line sensors;
Means for generating a processing unit array including a pixel of interest for image processing with respect to the reference color signal ;
Means for setting a correlation array of reference colors including the target pixel and having a smaller number of array elements than the processing unit array;
Means for extracting a correlation array group of color colors included in the scanning range in the processing unit array with the same number of array elements as the correlation array of the reference color
Means for calculating a correlation between the correlation array of the reference colors and each correlation array included in the correlation array group of the color colors;
Means for calculating a color color correlation array having the highest correlation with the reference color correlation array;
Means for obtaining a regression line from the pixel values included in the image correlation array of the color color having the highest correlation with the pixel values included in the correlation array of the reference color;
Means for generating a color pixel value for the pixel of interest using the regression line;
An image processing apparatus.   The image processing apparatus according to claim 1, wherein a correlation coefficient is used for the correlation.   5. The method according to claim 1, wherein when the line sensor unit does not include a monochrome line sensor, an addition average of signals from a color line sensor R, a line sensor G, and a line sensor B is used as the reference color signal. 2. The image processing apparatus according to 2.   3. The image processing apparatus according to claim 2, wherein when the line sensor unit includes a monochrome line sensor, an output signal from the monochrome line sensor is used as the reference color signal. When the monochrome line sensor is higher than the resolution of the color line sensor, the output of the monochrome line sensor is temporarily set to be the same as the resolution of the color line sensor. A low resolution conversion unit that performs low resolution processing and generates the reference color signal ;
5. The image processing according to claim 4, wherein high resolution conversion is performed for newly generating a pixel value of a color signal by using the regression line for a pixel value of a monochrome signal included in the target pixel of the reference color. apparatus. Obtaining an output signal of a line sensor unit composed of a plurality of line sensors including at least a color line sensor R, a line sensor G, and a line sensor B in units of pixels;
Generating a reference color signal using one or more of the output signals of the plurality of line sensors;
Generating a processing unit array including a target pixel of image processing for the reference color signal ;
A step of setting a correlation array of reference colors including the target pixel and having a smaller number of array elements than the processing unit array; and color colors included in the scanning range in the processing unit array with the same number of array elements as the correlation array of the reference colors Extracting a correlation sequence group of
Calculating a correlation between the correlation array of the reference color and each correlation array included in the correlation array group of the color color;
Calculating a color color correlation array having the highest correlation with the reference color correlation array;
Obtaining a regression line from the pixel values included in the correlation array of the reference color and the pixel values included in the image correlation array of the color color having the highest correlation;
Generating a color pixel value for the pixel of interest using the regression line;
An image processing method comprising:   When the line sensor unit does not include a monochrome line sensor, the method includes a step of calculating an average of signals of color line sensors R, line sensors G, and line sensors B as the reference color signal. The image processing method according to claim 6.   8. The image processing method according to claim 7, further comprising a step of using the reference color signal as an output signal from a monochrome line sensor when the line sensor unit includes a monochrome line sensor. When the monochrome line sensor is higher than the resolution of the color line sensor, the output of the monochrome line sensor is temporarily set to be the same as the resolution of the color line sensor. Performing low resolution processing to generate the reference color signal ;
Performing a high-resolution conversion for newly generating a high-resolution pixel value of a color signal using the regression line for a pixel value of an output signal of the monochrome line sensor included in the target pixel of the reference color The image processing method according to claim 8.

Images