JP2004304637A - Image processing method and image processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method and an image processing apparatus capable of properly carrying out the white board contrast emphasis processing. <P>SOLUTION: An image is divided into a plurality of blocks each comprising a plurality of pixels, a group of blocks comprising a prescribed number of blocks respectively longitudinally and laterally wherein each of all the blocks includes a target block, is set, and each target block is decided to be located at a corner remotest from a center O of the image in each block of the block group. Then a prescribed number of blocks with low luminance is extracted from the block group, an average Yab of a block background level of the blocks is calculated and the average Yab is set to the block background level of the target block. The processing above is sequentially carried out from the block located in the middle of the image toward the blocks located at the peripheral of the image. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image processing method and an image processing apparatus for performing a background removal process on an image obtained by photographing a white board or the like.
[0002]
[Prior art]
In general, a digital camera can freely control the image quality of an image captured by image processing. Therefore, by performing appropriate processing on the image quality of the captured image according to the purpose of shooting and the type of subject, a silver halide film can be used. This is advantageous in that an image with more favorable image quality can be obtained as compared with a camera that shoots images at a higher speed. For this reason, it is used not only for normal photographing but also for photographing, for example, characters and figures drawn on a whiteboard in a conference hall.
[0003]
When photographing a whiteboard on which characters, figures, and the like are drawn by a digital camera, the purpose of the photographing is to record information such as characters and figures on the whiteboard. For this reason, it is necessary to improve the clarity of information portions such as characters and figures by performing a so-called background removal process for ensuring that the white background portion of the whiteboard is pure white for the captured image as described above. Can be
[0004]
Japanese Patent Application Laid-Open Publication No. H11-163873 discloses that, when photographing a whiteboard, an object light image is converted into three primary colors in order to improve the clarity of information portions such as characters and figures by skipping the background (whiteboard portion) of the whiteboard. Γ characteristic for converting a pixel signal of a predetermined white background level or higher into a white saturation level using the pixel signal of each color component for each color component image, and using this γ characteristic A configuration configured to perform the correction is disclosed.
[0005]
[Patent Document]
JP-A-10-210287 [0006]
[Problems to be solved by the invention]
By the way, when the white board reflects, for example, the light of an electric lamp, an image obtained by imaging the white board has uneven illuminance and is hard to see.
[0007]
In the above-mentioned patent document, in order to reduce the illuminance unevenness, a captured image is divided into a plurality of blocks, and a white background level of each block is determined using a histogram of a level distribution of pixel data included in each block. The level is set as the white background level of the center pixel of the block, and the white background level of the pixel between the center pixels of the adjacent blocks is determined by linear interpolation using the white background levels of these center pixels.
[0008]
However, when determining the white background level of the pixel between the center pixels of the block, if the white background level of the block obtained by using the histogram of the level distribution of the pixel data included in each block as described above is used as it is, In some cases, the white background level of a pixel located around the pixel where the reflected light is imaged is different from the original white background level. In particular, for a pixel close to the position of the pixel that captured the reflected light of the electric light, the white background level calculated by linear interpolation has a large percentage of the white background level of the pixel that has captured the reflected light of the electric light. It often differs greatly from the actual white background level, and it has been difficult to properly perform the background removal processing.
[0009]
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an image processing method and an image processing apparatus that can appropriately perform background removal processing.
[0010]
[Means for Solving the Problems]
According to the first aspect of the present invention, an image is divided into a plurality of blocks including a plurality of pixels, a background level is set for each of the divided blocks, and a background level of a predetermined block in a block group including a plurality of blocks is set. Performing a smoothing process on the basis of the average value of the background level in a predetermined number of blocks among the other blocks in the block group, and using the background level calculated by the smoothing process, the center pixel and the center of each block. An image processing method comprising calculating a background level of a pixel located between pixels.
[0011]
According to a fifth aspect of the present invention, there is provided a dividing unit that divides an image into a plurality of blocks including a plurality of pixels, a setting unit that sets a background level for each of the divided blocks, and a predetermined unit in a block group including a plurality of blocks. Means for performing a smoothing process for setting the background level of the block to an average value of the background levels in a predetermined number of blocks among the other blocks in the block group; and calculating the background level calculated by the smoothing process. An image processing apparatus comprising: a calculating unit configured to calculate a base pixel of a central pixel of each block and a pixel located between the central pixels.
[0012]
According to the first or fifth aspect of the present invention, a background level is set for each of blocks obtained by dividing an image into a plurality of blocks including a plurality of pixels, and the background level of a predetermined block in a block group including a plurality of blocks is set. Is set to the average value of the background level in a predetermined number of blocks among the other blocks in the block group, so that the background level of the block having the pixel in which the reflected light or the like of the electric light is imaged is obtained. In addition, the difference from the background level of another block that does not image the reflected light of the electric light or the like is reduced.
[0013]
Thereby, when calculating the background level of the center pixel of each block and the pixels located between the center pixels using the background level calculated by the smoothing process, the center pixel of each block and the space between the center pixels are calculated. It is possible to avoid that the background level of the located pixel is significantly different from the original background level, and as a result, it is possible to perform appropriate background removal processing.
[0014]
According to a second aspect of the present invention, in the image processing method according to the first aspect, the predetermined block is located at a position farthest from the center of the image among blocks in a block group to which the predetermined block belongs. It is characterized by being.
[0015]
According to this invention, since the predetermined block in the block group is set to be located farthest from the center of the image among the blocks of the block group to which the predetermined block belongs, the reflected light of the electric light is The background level is calculated in a block closer to the center of the image in which is unlikely to appear. Thus, the background level of each pixel can be set more appropriately.
[0016]
According to a third aspect of the present invention, in the image processing method according to the first or second aspect, the smoothing process is performed in order from a block group located at the center of the image to a block group located on the peripheral side. It is characterized by the following.
[0017]
According to the present invention, since the smoothing process is performed in order from the block group located at the center of the image to the block group located on the peripheral side, there is little possibility that reflected light of electric light or the like is captured. The smoothing process is performed from the pixel located at the center of the image. This makes it possible to more appropriately set the background level of the pixel.
[0018]
According to a fourth aspect of the present invention, in the image processing method according to any one of the first to third aspects, the smoothing process includes setting a background level of the predetermined block to a background level of each block in the block group. Among them, the average value of the background levels of a plurality of blocks having a lower luminance is used.
[0019]
According to the present invention, in the smoothing processing, the background level of the predetermined block is set to an average value of the background levels of a plurality of blocks having lower luminance among the background levels of the blocks in the block group. Therefore, even if there is a block in which an image of the reflected light of the electric light or the like is present in the block group, it is possible to avoid performing the smoothing process using this block as much as possible. As a result, the background level of the pixel can be set more appropriately.
[0020]
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment of the present invention will be described.
[0021]
FIG. 1 is a control block diagram of a digital camera employing the image processing method according to the present invention.
[0022]
As shown in FIG. 1, the digital camera 1 includes an optical system 2, an image sensor 3, a signal processing unit 4, an A / D conversion unit 5, an image processing unit 6, an image memory 7, a control unit 8, a storage unit 22, and a VRAM 23. , A display unit 24 and an operation unit 25.
[0023]
The optical system 2 includes a zoom lens driven in the direction of the optical axis, and forms an optical image of a subject on an image plane of the image sensor 3.
[0024]
The image pickup device 3 includes a plurality of photoelectric conversion elements such as photodiodes, which are two-dimensionally arranged in a matrix, and R (red), G (green), and B (blue) colors are provided on a light receiving surface of each photoelectric conversion element. This is a CCD color area sensor in which filters are arranged at a ratio of 1: 2: 1.
[0025]
The image sensor 3 converts the optical image of the subject formed by the optical system 2 into electrical signals (images) of R (red), G (green), and B (blue) color components every 1/30 (second), for example. ), And output as image signals of R, G, and B colors.
[0026]
The image pickup device 3 performs an image pickup operation such as start and end of an exposure operation of the image pickup device 3 and readout (horizontal synchronization, vertical synchronization, and transfer) of an output signal of each pixel in the image pickup device 3 by a timing generator or the like (not shown). Controlled.
[0027]
The signal processing unit 4 includes a CDS (correlated double sampling) circuit, an AGC (auto gain control) circuit, and the like. The analog image signal output from the image sensor 3 is used to reduce noise of the image signal by the CDS circuit. Then, predetermined analog signal processing such as adjusting the level of an image signal by an AGC circuit is performed.
[0028]
The A / D conversion unit 5 converts the analog R, G, and B image signals output by the signal processing unit 4 into digital value image signals (hereinafter, referred to as digital signals) including a plurality of bits. Things.
[0029]
The image processing unit 6 performs black level correction for correcting the black level to a reference black level for each of the R, G, and B digital signals A / D-converted by the A / D conversion unit 5, and a white level corresponding to the light source. White balance for performing level conversion of digital signals of each color component of R (red), G (green), and B (blue) based on a reference, digital signals of R (red), G (green), and B (blue) The gamma correction is performed to correct the gamma characteristic.
[0030]
The image memory 7 is a memory that temporarily stores an image output from the image processing unit 6 and is used as a work area for performing a process described later on the image data by the control unit 8. The image memory 7 is, for example, a RAM.
[0031]
The control unit 8 includes a microcomputer having a built-in storage unit (not shown) including, for example, a ROM for storing a control program and a RAM for temporarily storing data, and drives each unit in the digital camera 1 described above. In this way, the photographing operation of the digital camera 1 is comprehensively controlled.
[0032]
In the present embodiment, for example, when a white board or the like on which characters, figures, and the like are drawn is taken, the control unit 8 ensures that a white background portion (a portion on which characters and the like are not drawn) of the white board is taken. It is characterized in that it is made pure white and image processing for clarifying characters is performed.
[0033]
Here, the outline of this image processing will be described. Each of the R, G, B digital signals output from the image processing unit 6 is converted into a luminance signal Y and color difference signals Cr, Cb for the following processing convenience. I do. Then, the entire image is divided into a plurality of blocks, and a histogram for the luminance Y is generated for each block.
[0034]
Next, in the generated histogram, a maximum part (peak) of the frequency is searched for on a low luminance side and a high luminance side from a certain luminance (an average luminance Ya described later), and the base level (described later) is used for each block using this peak. Block background level). Next, after correcting (smoothing) the block base level, a base level (pixel base level described later) for each pixel is calculated from the corrected block base level.
[0035]
Next, after performing tone correction for each pixel using the above-described pixel base level, noise removal processing and connection processing of broken portions such as characters are performed, and the luminance of red, yellow, etc., not black, is compared. Tone correction is performed to make characters and the like written with high-quality ink easy to see.
[0036]
Then, after performing the processing for improving the saturation of the character area and the like and the outline correction processing, the luminance signal Y and the color difference signals Cr, Cb are converted into the original R, G, B digital signals, and the processing is terminated.
[0037]
As shown in FIG. 1, the control unit 8 includes an RGB-YCrCb conversion unit 9, a block setting unit 10, a histogram generation unit 11, a block base level setting unit 12, and a smoothing process for realizing the above-described image processing. Unit 13, pixel background level calculation unit 14, first gradation correction processing unit 15, noise removal / connection unit 16, first saturation correction unit 17, second gradation correction processing unit 18, second saturation correction unit 19 , A contour correction unit 20 and a YCrCb-RGB conversion unit 21.
[0038]
In the following description, on the surface of the whiteboard on which characters and the like are drawn, a portion where the characters and the like are drawn is referred to as a character region, and a white portion where nothing is drawn is referred to as a white plate region.
[0039]
The RGB-YCrCb conversion unit 9 converts the R, G, and B digital signals output from the image processing unit into a luminance signal Y and a color difference according to the following equations (1) to (3) in order to generate a luminance histogram. The signals are converted into signals Cr and Cb.
[0040]
Y = 0.299R + 0.587G + 0.114B (1)
Cr = 0.500R-0.419G-0.081B (2)
Cb = −0.169R−0.332G + 0.500B (3)
The block setting unit 10 divides the image after the conversion processing by the RGB-YCrCb conversion unit 9 into a predetermined number of blocks. For example, as illustrated in FIG. 2A, the block setting unit 10 captures an image of a whiteboard on which a character “ABC” is written on a plate surface by using the image sensor 3 and converts the whiteboard by the RGB-YCrCb conversion unit 9. Is divided into a plurality of blocks as shown in FIG.
[0041]
At this time, in the present embodiment, the block setting unit 10 generates a peripheral light drop of the lens so that, for example, only one portion where the peripheral light drop of the lens is generated constitutes one block as much as possible. Divide the size into as small as possible compared to the size of the range.
[0042]
This is because, as described above, the setting of the block base level in each block is performed by searching for a frequency peak from a certain luminance to a low luminance side and a high luminance side. If there is a mixture of a part where the light is generated and a part where the light is not generated, for example, even if the appropriate block base level of the block is the luminance level of the part where the peripheral light intensity drop occurs, This is because the luminance level of the portion where the light quantity drop does not occur may be set as the block base level of the block, and the block base level setting unit 12 may set an inappropriate block base level.
[0043]
In addition, the amount of light incident on the outer peripheral side of the lens is obliquely incident on the lens, so that the amount of light incident per unit area of the lens is incident on the central portion of the lens. This is a phenomenon in which light is less than light, and as a result, the periphery (particularly, four corners) of an image is dark.
[0044]
Also, as described above, the image after the conversion processing by the RGB-YCrCb conversion unit 9 is divided into a plurality of blocks because the variation in the character density (the ratio of the character portion to the white background portion) of the handwritten on the whiteboard is caused. In the case of photographing, the light source is not constant, and the illuminance distribution in the image is greatly different due to the uneven illuminance in the captured image. This is for setting the level.
[0045]
The histogram generation unit 11 generates a histogram of the luminance Y in the pixels of the block for each block generated by the block setting unit 10.
[0046]
FIG. 3 shows various histograms showing the frequency distribution of the luminance Y. The horizontal axis represents the luminance Y, and the vertical axis represents the number of pixels N.
[0047]
For example, the number of pixels that image a region of a character drawn with low-luminance ink (for example, black ink), a region of a white plate, a region of reflecting a light source (for example, light of a light), or a region of a metal frame of a whiteboard are compared. In the case of a block with many targets, as shown in FIG. 3A, the number of pixels is maximized at a relatively low luminance level in a character area (arrow A), and a relatively high luminance in a white board area is obtained, as shown in FIG. The number of pixels is maximized at the level (arrow B), and the number of pixels is maximized at a certain high brightness level in the area reflecting the light source and the area of the metal frame of the whiteboard (arrow C).
[0048]
Further, a histogram of a block having a relatively large number of pixels that image a region of an object having a very low luminance, a white plate region, a region reflecting a light source (for example, light of a lamp), and a region of a metal frame of a whiteboard is shown in FIG. As shown in FIG. 3 (b), the number of pixels is maximized at a very low luminance level in the region of the object (arrow D), and the number of pixels is maximized at a relatively high luminance level in the white board region (arrow D). E), the number of pixels is maximized at a certain high luminance level in the area reflecting the light source and the area of the metal frame of the whiteboard (arrow F).
[0049]
In addition, a region of a character drawn with low-luminance ink, a region that reflects a light source and that is higher than the luminance level of the peak indicated by the arrow C in FIG. 3A or the arrow F in FIG. In the histogram of a block having a relatively large number of imaged pixels, for example, as shown in FIG. 3C, the number of pixels is maximized at a relatively low luminance level of a character area (arrow G), and the light source is reflected. At a very high luminance level in the area and in the area of the metal frame of the whiteboard, the number of pixels is maximal (arrow H).
[0050]
Further, in addition to the area of characters drawn with low-luminance ink (for example, black ink), the area of a white plate, the area of reflecting a light source (for example, light of an electric light), and the area of a metal frame of a white board, for example, For a block having pixels obtained by imaging a character region whose luminance is slightly similar to a white plate region, such as a character written with ink such as yellow, if the histogram is faithfully created, for example, as shown in FIG. The histogram shown in FIG. 3A has a peak (arrow I) at a luminance level slightly lower than the luminance level of the white plate area.
[0051]
However, as described above, since the setting of the block base level in each block is performed by searching for a frequency peak from a certain luminance to a low luminance side and a high luminance side, it is written with ink such as orange. When the brightness of the area such as a character becomes a value between the brightness of the white board area and the brightness of the entire image described later, the place where the brightness of the white board area should be set as the block base level is written with orange ink. There is a possibility that the luminance of the area of the drawn character or the like is erroneously set as the block base level.
[0052]
Therefore, in the present embodiment, as shown in FIG. 3 (e), the value of the color difference signals Cr and Cb is equal to or less than a predetermined threshold value, that is, a histogram is created using only pixel data with low saturation, so that a white board is created. The peak in the luminance level of the character area similar to the area is removed in advance.
[0053]
Assuming that the background of the object to be photographed (the whiteboard area of the whiteboard in the above example) is referred to as a background, the block background level setting unit 12 uses the histogram created for each block to set the background of each block. This is for setting the luminance level. Note that the luminance level of the background in the block set in this way is called a block background level.
[0054]
The block base level setting unit 12 first calculates the average luminance of the surface of the whiteboard. Here, since it is considered that the whiteboard is likely to exist in the center of the image, the block base level setting unit 12 sets the average luminance of the surface of the whiteboard as shown in FIG. In order to calculate accurately, the average luminance Ya is calculated from the central part of the image (for example, the central part of 60% in the vertical and horizontal directions of the entire image).
[0055]
Then, in the various histograms as described above, a luminance level closest to the average luminance Ya of the calculated image among the luminance levels having a plurality of local maxima is set as the block base level.
[0056]
For example, assuming that the histogram for a certain block is as shown in FIG. 4A and that the calculated average luminance Ya of the image is the luminance shown in FIG. 4A, first, the average luminance Ya , The magnitude of the frequency value (frequency) at each luminance level is determined from the high luminance side (the direction of arrow P). Then, a peak Y _H peaks first appeared.
[0057]
The detection of the peak Y _H, for example, a certain luminance level of frequency values, or higher frequency value than the plurality of luminance levels different high luminance side whether, and the low luminance side of a plurality of different luminance levels It is determined whether or not the frequency value is higher than the frequency value of the luminance level to be compared, or if the frequency value is higher than or equal to the frequency value of the luminance level to be compared, it is determined to be a peak.
[0058]
Here, if a very high luminance level (arrow J) as shown in FIG. 4B is set as the block base level, the following problem occurs.
[0059]
That is, the pixel background level calculation unit 14, which will be described later, considers the block background level of each block as the background level of the pixel located at the center position of the block, and determines the background level of the pixels located between the center pixels of each block. When a high luminance level is set as the block base level of the block to be calculated by the linear interpolation, the base level of the pixels between the center pixels calculated by the pixel base level calculation unit 14, particularly the pixels located near the center pixel, is calculated. May be significantly different from the block base level of the block to which the pixel belongs.
[0060]
In the present embodiment, in order to prevent such a problem, as shown in FIG. 4C, it is assumed that the upper limit value Ymax (the luminance level is expressed by 256 gradations) smaller than the above-described extremely high luminance level. For example, 250 gradations) are set, and the magnitude of the frequency value is determined for the luminance level up to the upper limit Ymax.
[0061]
Accordingly, even if a block includes a pixel that captures an object having an extremely high luminance (in this embodiment, a luminance whose gradation exceeds 250), such luminance is set as a block base level. Setting can be prevented.
[0062]
On the other hand, in the case of a block having a histogram as shown in FIG. 4B, since there is no peak between the average luminance Ya and the upper limit Ymax, the magnitude of the frequency value is determined up to the upper limit Ymax as described above. Is performed, no peak is detected. Further, for example, even when all the pixels in the block have luminance lower than the average luminance Ya, no peak is detected.
[0063]
In this case, by setting the average luminance Ya as the block base level of the block, a luminance level close to the luminance level of the background (white background) can be set as the block base level.
[0064]
Next, in the histogram shown in FIG. 4A, as shown in FIG. 4D, the magnitude of the frequency value at each luminance level is determined from the average luminance Ya to the lower luminance side (the direction of arrow Q). Go. At this time, a peak Y _L peaks first appeared.
[0065]
The detection of the peak Y _L, as well as the detection of the peak Y _H, for example, the frequency value of a certain brightness level, whether the high frequency value than the plurality of luminance levels different high luminance side and low luminance It is determined whether or not the frequency value is higher than a plurality of luminance levels different from each other, and if the frequency value is higher than or equal to the frequency value of the luminance level to be compared, the peak is determined.
[0066]
Further, as shown in FIG. 4E, when a very low luminance level such as a magnet attached to a whiteboard or a very dark background is set as the block base level, the following problem occurs. .
[0067]
That is, in the same manner as described above, the pixel background level calculation unit 14, which will be described later, regards the block background level of each block as the background level of the pixel located at the center position of the block, and determines the pixels located between the center pixels of each block. When the low luminance level is set as the block base level of the block, the base level of the pixel between the center pixels calculated by the pixel base level calculation unit 14 is determined by the linear interpolation. In some cases, the level of the block greatly differs from the block base level.
[0068]
In this embodiment, in order to prevent such a problem, as shown in FIG. 4F, it is assumed that the lower limit value Ymin (the luminance level is expressed by 256 gradations), which is larger than the extremely low luminance level described above. (For example, 10 tones), and the magnitude of the frequency value is determined for the luminance level up to the lower limit Ymin.
[0069]
On the other hand, in the case of a block having a histogram as shown in FIG. 4E, since there is no peak between the average luminance Ya and the lower limit Ymin, the magnitude of the frequency value is determined up to the lower limit Ymin as described above. However, no peak is detected. Further, for example, even when all the pixels in the block have a luminance higher than the average luminance Ya, no peak is detected.
[0070]
In this case, by setting the average luminance Ya as the block base level of the block, a luminance level close to the background luminance level can be set as the peak level.
[0071]
As described above, when detecting the peak Y _L peak Y _H and the low-luminance side of the high-luminance side, the closer to either the average luminance Ya, set as a block ground level Yb of the block. For example, in FIG.
_{Ya-Y L> Y H -Ya} ... (4)
Since it is, to set the peak Y _H of the high luminance side as a block ground level Yb of the block.
[0072]
By the way, when the brightness of the light source reflected on the whiteboard, the area of the metal frame of the whiteboard, or the area of the magnet attached to the whiteboard is a value between the upper limit value Ymax and the lower limit value Ymin. However, in a block in which these regions are imaged, if the block base level is set based on only the histogram of the luminance of the block, the high luminance level or the low luminance level is set as the block base level of the block. Will be set.
[0073]
However, in the same manner as described above, the pixel background level calculation unit 14, which will be described later, regards the block background level of each block as the background level of the pixel located at the center position of the block, and determines the pixels located between the center pixels of each block. In order to calculate the background level of the block by linear interpolation, if a high brightness level or a low brightness level is set as the block base level of the block, the base level of the pixel between the central pixels calculated by the pixel base level calculation unit 14 is In some cases, the level greatly differs from the block base level of the block to which the pixel belongs.
[0074]
In order to avoid this problem, the smoothing processing unit 13 performs the processing for each of the blocks set by the block background level setting unit 12 in order from the block located at the center of the image to the blocks located at the periphery for all the pixels. This is for smoothing the block base level (reducing the difference between the block base levels of adjacent blocks).
[0075]
Performing the smoothing process in order from the block located at the center of the image to the blocks located at the periphery is because the smoothing process is performed from the pixel located at the center of the image where the reflected light of the electric light is unlikely to appear. By doing so, a more appropriate pixel background level is set.
[0076]
FIG. 5 is a diagram for explaining the smoothing process performed by the smoothing processing unit 13.
[0077]
As shown in FIG. 5, when performing the smoothing process on a block to be smoothed (hereinafter, referred to as a target block, which is indicated by oblique lines in FIG. 5), the smoothing processing unit 13 includes the target block. , A block group including a predetermined number of blocks in each of the vertical and horizontal directions (in the present embodiment, three blocks in each of the vertical and horizontal directions).
[0078]
At this time, a block used for the smoothing (hereinafter, referred to as a smoothing block) is determined such that the target block is located at a corner farthest from the center O of the image. For example, as shown in FIG. 5, the entire image is divided into two equal parts in the vertical and horizontal directions by a line segment including the center O of the image, and when the target block is located in the upper left area T1, the target block is located in the upper left corner. Determine the smoothing block.
[0079]
Similarly, when the target block is located in the lower left area T2, the smoothed block is determined so that the target block is located in the lower left corner, and when the target block is located in the upper right area T3, the smoothed block is located in the upper right corner. When the block is located in the lower right area T4, the smoothed block is determined so that the target block is located in the lower right corner.
[0080]
FIG. 5 is a diagram in which a set of a target block and a smoothed block in each of the regions T1 to T4 are extracted and drawn.
[0081]
The reason for setting the smoothing block in this way is that the area where the reflection of the light source or the metal frame of the whiteboard exists is often located at the periphery of the image, while the area closer to the center of the screen is affected by the light source etc. Since there is a high possibility that there is an area that has not been affected by light, the block on the center side of the screen, which is not affected by the light source, is used for smoothing the block of interest. This is because the difference between the target block and the block base level is reduced, and the smoothing can be performed appropriately.
[0082]
The smoothing processing unit 13 extracts a predetermined number (three in this embodiment) of low-brightness blocks from the smoothed block and the target block set in this way, and calculates the block base level of these blocks. The average Yab is calculated, and the average Yab is set to the block base level of the target block.
[0083]
Further, as shown in FIG. 5B, in order to set a block base level appropriately by performing a smoothing process from a block located at the center of an image in which the reflected light of the light source is unlikely to be reflected, for example, When the blocks in the upper right area T3 are numbered, the smoothing processing unit 13 performs, for example, 1 → 2 → 6 → 5 → 9 → 10 → 11 → 7 → 3 → 4 → 8 → 12 →. → 2 → 6 → 5 → 3 → 7 → 11 → 10 → 9 → 4 → 8 → 12 → ... or 1 → 2 → 5 → 3 → 6 → 9 → 4 → 7 → 10 → 8 → 11 From the block located at the center of the image to the periphery, such as → 12 ..., 1 → 2 → 5 → 9 → 6 → 3 → 4 → 7 → 10 → 8 → 11 → 12 ... The smoothing process is performed in order toward the located block.
[0084]
The pixel base level calculation unit 14 sets the block base level of each block subjected to the smoothing processing by the smoothing processing unit 13 to the base level of a pixel located at the center position of the corresponding block (hereinafter, referred to as a center pixel). Then, the background level of the pixels located between the center pixels (hereinafter, referred to as pixel background level) is calculated by linear interpolation.
[0085]
In FIG. 6, it is assumed that one cell represents one block, the pixel base level of the center pixel G1 of the block B1 is Y1, the pixel base level of the center pixel G2 of the block B2 is Y2, and the center pixel G3 of the block B3 is a pixel. the ground level Y3, when the pixel ground level of the center pixel G4 block B4 Y4, for example a pixel background level _{Y G} of an arbitrary pixel Gx in the block B1, is calculated by these pixels background level Y1 to Y4.
[0086]
That is, the pixel ground level Y _G of the pixel Gx is calculated by the following equation (5).
Y _G = (α · Y1 + β · Y2 + γ · Y3 + δ · Y4) / 4 (5)
It is calculated by: Note that α, β, γ, and δ are weighting factors, and each weighting factor has a relationship proportional to the distance from the pixel Gx to the center pixel G1 to G4 of each of the blocks B1 to B4.
[0087]
The first gradation correction processing unit 15 sets a correction function as shown in FIG. 7 for each pixel using the pixel background level calculated by the pixel background level calculation unit 14, and based on the correction function, The key correction is performed.
[0088]
FIG. 7 shows the correction function of the gradation correction by the first gradation correction processing unit 15, and the horizontal axis indicates the input (the pixel of the image after the RGB-YCrCb conversion processing) of the first gradation correction processing unit 15. The vertical axis represents the output of the first gradation correction processing unit 15.
[0089]
In the case of the correction function shown in FIG. 7, the output is set to 0 when the input is less than 100, and the output is set to 255 when the input is equal to or higher than the pixel base level Yβ calculated by the pixel base level calculator 14. Further, when the input level is between 100 and the above-described pixel base level Yβ, the input and output are set to have a linear relationship, and gradation correction is performed according to this linear function.
[0090]
According to the gradation correction, the brightness level of each pixel in the image after the RGB-YCrCb conversion process is equal to or higher than the pixel base level Yβ calculated by the pixel base level calculation unit 14 of the pixel. Is converted to the highest luminance (Y = 255 in the present embodiment), so that the area to be set as the background (white plate) is set to pure white.
[0091]
Further, the luminance level Y _Q represented by the following equation (6)
_YQ = {25500 / (355-Yβ)} (6)
And a pixel having a luminance level smaller than the value _YQ is converted into a value smaller than the input level, so that the luminance level is smaller than the value of _YQ shown in the above equation (6). The contrast of the character area having the level with the background is increased.
[0092]
By the way, when the image processed by the image processing unit 6 includes small point-like noise, the pixel in the portion where the noise is generated is the pixel in the case where the noise is not generated. Has a luminance level lower than the pixel base level. Therefore, such a pixel cannot be converted to pure white (gradation is 255) even if gradation correction is performed by the correction function shown in FIG.
[0093]
For example, as shown in FIG. 7, in a region where the input level is equal to or higher than the pixel base level Yβ calculated by the pixel base level calculation unit 14, the first brightness correction processing unit 15 performs the gradation correction to obtain the highest luminance. (Gray level 255), a part of the character in the image after the RGB-YCrCb conversion processing has an area (e.g., a thin ink area) If there is a blurred area, the handwriting will be interrupted at this part.
[0094]
The noise removing / connecting unit 16 removes such noise and corrects it to the original background color (pure white), and also connects the part where the handwriting is interrupted (disconnected part).
[0095]
FIG. 8 is a diagram for explaining the noise removal / connection process performed by the noise removal / connection unit 16.
[0096]
For example, as shown in FIG. 8A, when it is assumed that noise or a discontinuous portion exists, an average value of luminance levels of all pixels and eight pixels located around the pixel is calculated, and The luminance level of the pixel is set to the average luminance level (hereinafter, this processing is referred to as average value processing). As a result, as shown in FIG. 8B, the image after the average value processing has blurred characters and noise.
[0097]
Then, the noise removing / connecting unit 16 determines whether or not the pixels having the luminance level which is not the highest luminance in the image after the average value processing are continuous, for example, by a predetermined number or more in a predetermined direction. If the number is smaller than the predetermined number, the pixel is determined as an isolated point, that is, noise, and the luminance level is converted to the maximum luminance. Thus, as shown in FIG. 8C, noise can be removed (hereinafter, this process is referred to as an isolated point removal process).
[0098]
In addition, in order to connect the broken portions of the character and clarify the blurred character, the noise removing / connecting unit 16 performs the image processing after the first gradation correction processing and the image after the average value processing and the isolated point removing processing. And the following processing is performed.
[0099]
In other words, the noise removing / connecting unit 16 images the white board region of the pixel having the highest luminance in both the image after the first gradation correction processing and the image after the average value processing and the isolated point removing processing. Is determined, and the luminance level of the pixel is set to the maximum luminance (pure white).
[0100]
In addition, the noise removing / connecting unit 16 has the highest luminance in the image after the first gradation correction processing, and processes the pixel that has become the highest luminance after the average value processing and the removal of the isolated point as follows. .
[0101]
Pixels that have the highest luminance in the image after the first gradation correction processing and no longer have the highest luminance after the average value processing and the removal of the isolated points have handwriting due to the above gradation correction, which should be a character area. The pixel in the interrupted portion and the pixel which had the highest luminance in the image after the first gradation correction processing but did not have the highest luminance after the average value processing and which should not be a character area (for example, FIG. b) pixels in the area indicated by the arrow L).
[0102]
Therefore, in the present embodiment, as shown in FIG. 8 (e), one of the two pixel rows that are continuous in the predetermined direction for the predetermined number or more is placed on an extension of the other pixel row. By judging whether or not there is, the broken portion and the direction of the handwriting (arrow K) in the broken portion are detected from the image after the first gradation correction processing, and the pixels (FIG. 8 (e) is detected.
[0103]
Then, regarding the pixels in the area to be the character area, the handwriting of the character is interrupted by the above-described gradation correction, and it is determined that the interrupted portion is connected by the average value processing, and the surrounding area considered as the character area is determined. The luminance level is calculated using the pixels. That is, the average value of the luminance levels of the eight pixels located around the pixel is set as the luminance level. Thereby, the disconnected portions are connected.
[0104]
On the other hand, the brightness level of the pixel which should not be the above-described character area is set to the highest brightness (pure white).
[0105]
Further, the noise removal / connection unit 16 does not have the highest luminance in the image after the first gradation correction processing, and generates noise (isolated point) for the pixel having the highest luminance after the average value processing and the isolated point removal processing. ), And it is determined that the noise has been removed by the isolated point removal processing, and the luminance level of the pixel is maintained at the maximum luminance converted in the isolated point removal processing.
[0106]
In addition, the noise removal / connection unit 16 determines that the image after the first gradation correction process and the pixel whose brightness level after the average value process / isolated point removal process is not the highest brightness are both images of the character region. It is determined, and the luminance level of the pixel is set to the luminance level of the pixel in the image after the first gradation correction processing.
[0107]
As a result, as shown in FIG. 8 (d), an image is obtained in which noise is removed from FIG. 8 (a), broken portions of the characters are connected, and blurred characters are clarified.
[0108]
The first chroma correction unit 17 generates the chrominance signals Cr and Cb so that the pixel whose luminance after the processing by the noise removal / connection unit 16 is converted to the highest luminance (gradation 255) does not have chroma. Is corrected to make the value of 0 zero.
[0109]
The second gradation correction processing unit 18 determines that a pixel having no highest luminance is a pixel in which a character area is imaged, and performs gradation correction using a correction function as shown in FIG.
[0110]
In FIG. 9, when the input level is 128 or less, the output level is set to be the same as the input level, and the input level is set to a value lower than 128 to make it easier to see light-colored characters such as characters written in orange ink. When the input image level is 255 or less, the slope of the correction function is made gentler than the correction function when the input level is 128 or less, so that an image captured by a pixel having a relatively high luminance exceeding the gray scale 128 can be darkened. Is corrected. Note that this gradation correction is performed with particular consideration to make the above-described characters easier to see when an image on which various processes have been completed by the respective units of the control unit 8 is printed by a predetermined image forming apparatus. .
[0111]
The second saturation correction unit 19 applies a predetermined gain to the color difference signals Cr and Cb on the character area after gradation correction to make it easier to see the characters and the like written on the whiteboard, and adjusts the saturation. Is to increase.
[0112]
The contour correction unit 20 performs a contour correction for enhancing a contour of a character or the like by a known general method.
[0113]
The YCrCb-RGB conversion unit 21 converts the luminance signal Y and the chrominance signals Cr and Cb subjected to the contour correction by the contour correction unit 20 into RGB image signals according to the following equations (7) to (9). It is.
[0114]
R = Y + 1.402Cr (7)
G = Y-0.714Cr-0.344Cb (8)
B = Y + 1.772Cb (9)
The storage unit 22 stores the image on which the various processes described above in the control unit 8 have been performed.
[0115]
The VRAM 23 is a buffer memory of an image displayed on the display unit 24, and has a storage capacity of an image corresponding to the number of pixels of the display unit 24.
[0116]
The display unit 24 displays an image that has been subjected to the various processes described above in the control unit 8 and temporarily stored in the VRAM 23.
[0117]
The operation unit 25 switches an operation unit for switching between a mode in which the control unit 8 executes the above-described various processes and a mode in which a normal imaging operation is performed without performing these processes in order to shoot a whiteboard or the like. Including.
[0118]
Next, the procedure of the image processing method of the present invention will be described.
[0119]
FIG. 10 is a flowchart showing the procedure of the image processing method of the present invention.
[0120]
As shown in FIG. 10, first, each of the R, G, and B digital signals is output by the image processing unit 6 (step # 1), and the RGB-YCrCb conversion unit 9 converts these digital signals into the luminance signal Y. After being converted into the color difference signals Cr and Cb (step # 2), the entire image is divided into a plurality of blocks by the block setting unit 10 (step # 3), and the histogram generation unit 11 outputs the pixels of the blocks for each block. Is generated (step # 4).
[0121]
Then, the block base level setting unit 12 sets the block base level in each block using the generated histogram (step # 5), and the smoothing processing unit 13 smoothes the block base level (step # 5). 6) Thereafter, the pixel base level calculation unit 14 calculates a pixel base level for each pixel from the block base level after the smoothing process (step # 7).
[0122]
Next, the first gradation correction processing unit 15 performs gradation correction for each pixel using the above-described pixel base level (step # 8), and the noise removal / connection unit 16 performs noise removal processing and character Then, the connection processing of the disconnection part such as is performed (step # 9).
[0123]
If the luminance of the pixel after processing by the noise removal / connection unit 16 does not have the maximum luminance (YES in step # 10), the second gradation correction processing unit 18 compares the luminance of the pixel with the luminance of FIG. After the gradation correction is performed by the correction function shown in (1) (Step # 11), the second saturation correction section 19 performs processing for improving the saturation of the character area and the like (Step # 12). On the other hand, in step # 10, if the luminance of the pixel after the processing by the noise removing / connecting unit 16 has the highest luminance (NO in step # 10), the first chroma correction unit uses the color difference signal for the pixel. Correction is performed to set the values of Cr and Cb to 0 (step # 13).
[0124]
Then, when the processing of step # 12 or # 13 is completed, the contour correction processing is performed by the contour correction section 20 (step # 14), and then the luminance signal Y and the color difference signals Cr, Cb are converted by the YCrCb-RGB conversion section 21. The digital signals are converted into R, G, and B digital signals (step # 15).
[0125]
FIG. 11 is a flowchart showing a detailed procedure of the block base level setting process by the block base level setting unit 12 in step # 5 of FIG.
[0126]
As shown in FIG. 11, the average luminance Ya of the surface of the whiteboard is calculated (step # 21), and the magnitude of the frequency value at each luminance level is determined from the average luminance Ya to the upper limit Ymax on the higher luminance side. (Step # 22).
[0127]
Then, it is determined whether a peak is present (step # 23), when the peak is present (YES in step # 23), the first detected peak Y _H is set as the peak of the high-luminance side If no peak exists (step # 24) (NO in step # 23), average luminance Ya is set as a peak on the high luminance side (step # 25).
[0128]
Next, the magnitude of the frequency value at each luminance level is determined from the average luminance Ya to the lower luminance side to the lower limit Ymin (step # 26), and it is determined whether or not a peak exists (step # 27). .
[0129]
As a result, when the peak is present (YES in step # 27), if the first detected peak Y _L is set as the peak on the low luminance side (step # 28), there is no peak (step # 27 In (NO), the average luminance Ya is set as the peak on the low luminance side (step # 29).
[0130]
Then, the peak Y _L peak Y _H and the low-luminance side of the high-luminance side, which is set in step ♯24 and # 28, closer to either the average luminance Ya is set as a block ground level Yb of the block (Step # 30).
[0131]
Until the above processing is completed for all the blocks (NO in step # 31), the blocks to be processed are sequentially changed (step # 32), and the processing of steps # 22 to # 30 is performed, and all the blocks are processed. When the block background level setting process is completed (YES in step # 31), the process by the block background level setting unit 12 ends.
[0132]
FIG. 12 is a flowchart showing a detailed procedure of the smoothing process by the smoothing processing unit 13 in step # 6 in FIG.
[0133]
As shown in FIG. 12, first, when the target block is located in the upper left area of the image (YES in step # 41), a smoothing block is determined so that the target block is located in the upper left corner (step # 4). 42), when the target block is located in the lower left area of the image (YES in step # 43), a smoothing block is determined so that the target block is located in the lower left corner (step # 44), and the target block is determined. When the block is located in the upper right area of the image (YES in step # 45), the smoothed block is determined so that the target block is located in the upper right corner (step # 46), and the target block is located in the lower right area. At this time (NO in step # 45), a smoothing block is determined so that the target block is located at the lower right corner (step # 47).
[0134]
After extracting three blocks with low luminance from the smoothed block and the target block determined in any of steps # 41, # 43, # 45, and # 47 (step # 48), these blocks are extracted. Is calculated (step # 49), and the average value Yab is set as the block background level of the target block (step # 50).
[0135]
Until the above processing is completed for all the blocks (NO in step # 51), the blocks to be processed are sequentially changed (step # 52), and the processing of steps # 41 to # 50 is performed, and all the blocks are processed. When the smoothing process is completed for () (YES in step # 51), the process by the smoothing processing unit 13 ends.
[0136]
FIG. 13 is a flowchart showing a detailed procedure of the noise removal / connection process by the noise removal / connection unit 16 in step # 9 in FIG.
[0137]
As shown in FIG. 13, for all pixels, an average value processing is performed to calculate the average value of the luminance levels of eight pixels located around the pixel and set the luminance level of the pixel to the average luminance level. Is performed (step # 61), and it is determined whether or not the pixel whose luminance level is not the highest luminance is continuous in a predetermined direction for a predetermined number or more. If the number of continuous pixels is smaller than the predetermined number, Is performed to convert the luminance level of the pixel to the maximum luminance (step # 62).
[0138]
Then, in the image after the first gradation correction processing and the image after the average value processing and the isolated point removal processing, for the pixel having the highest luminance level (YES in steps # 63 and # 64), the value of the pixel is determined. The luminance level is set to the maximum luminance (step # 65).
[0139]
On the other hand, for the pixel having the highest luminance in the image after the first gradation correction processing (YES in step # 63), and no longer having the highest luminance after the average value processing and the removal of the isolated point (NO in step # 64), It is determined whether or not the pixel is to be a character area (step # 66). If the pixel is to be a character area (YES in step # 66), surrounding pixels considered to be a character area are used. The brightness level is calculated (step # 67), and if the pixel is not a character area (NO in step # 66), the brightness level of that pixel is set to the highest brightness (step # 68).
[0140]
Further, the image after the first gradation correction processing does not have the highest luminance (NO in step # 63), and has the highest luminance after the average value processing and the isolated point removal processing (YES in step # 69). Is that the luminance level of the pixel is maintained at the maximum luminance converted in the isolated point removal processing (step # 70), and the image after the first gradation correction processing and the luminance level after the average value processing and the isolated point removal processing are For a pixel having neither maximum luminance (NO in steps # 63 and # 69), the luminance level of the pixel is set to the luminance level of the pixel in the image after the first gradation correction processing (step # 71).
[0141]
As described above, since the block base level set for each block is smoothed using the smoothing block, the block base level of the block having the pixel that captures the reflected light of the light source and the like is determined based on the reflected light of the light source. Can be reduced from the block background level of other blocks that do not image the data.
[0142]
Thus, when calculating the center pixel of each block and the white background level of the pixel located between the center pixels using the block base level calculated by the smoothing process, the center pixel of each block and the space between the center pixels are calculated. Can be prevented from having a value that is significantly different from the original pixel background level, and as a result, appropriate background removal processing can be performed.
[0143]
In addition, the smoothing block is determined such that the block of interest is located at the corner farthest away from the center O of the image, and the block base with a block closer to the center of the image where the reflected light of the electric light is less likely to appear. Since the level is calculated, the pixel base level can be set more appropriately.
[0144]
In addition, by performing smoothing processing in order from the block group located at the center of the image to the block group located on the peripheral side, the pixel located at the center of the image where the reflected light of the electric light is unlikely to be reflected can be obtained. Since the smoothing process is performed, the pixel base level can be set more appropriately.
[0145]
Further, the block base level of the block of interest is set to be the average value of the block base levels of a plurality of blocks having lower luminance among the block base levels of the blocks in the block group to which the block of interest belongs. Even in the case where there is a block in which the reflected light of the electric light is captured in the group, it is possible to avoid performing the smoothing process using this block as much as possible, and to set the pixel base level more appropriately. Can be set to
[0146]
The present invention is not limited to the above embodiment, and the following modifications (1) and (2) can be adopted.
[0147]
(1) In the above embodiment, the RGB-YCrCb conversion unit 9, the block setting unit 10, the histogram generation unit 11, the block base level setting unit 12, the smoothing processing unit 13, the pixel base level calculation unit 14, the first gradation The correction processing unit 15, the noise removal / connection unit 16, the first saturation correction unit 17, the second tone correction processing unit 18, the second saturation correction unit 19, the contour correction unit 20, and the YCrCb-RGB conversion unit 21 Although the case of mounting on the camera 1 has been described, the present invention is not limited to this, and it may be configured to be mounted on the control unit 27 of a personal computer 26 (hereinafter, referred to as PC 26) as shown in FIG.
[0148]
In this case, as shown in FIG. 14, the PC 26 takes in an image subjected to processing such as γ correction from an image input device 28 such as a digital camera, and generates a histogram, sets a block base level, After performing the conversion process, the calculation of the pixel background level, the noise removal / connection process, and the like, the image data is stored in the image storage device 29 such as a hard disk device.
[0149]
When the image is to be output such as printed on paper, the PC 26 fetches the image from the image storage device 29, sends the image to the image output device 29 such as a printer, and sends the image to the image output device 30. You may make it print out. The operation unit 31 is used to input an instruction to the control unit 27 to execute a histogram generation process and the like, and indicates a mouse or a keyboard.
[0150]
(2) The PC 26 shown in FIG. 14 and the digital camera 1 in the above embodiment are functionally controlled by RGB-YCrCb conversion units 9 and 32, block setting units 10 and 33, histogram generation units 11 and 34 by a predetermined program. Block background level setting units 12 and 35, smoothing processing units 13 and 36, pixel background level calculation units 14 and 37, first gradation correction processing units 15 and 38, noise removal / connection units 16 and 39, first saturation Correction sections 17 and 40, second tone correction sections 18 and 41, second saturation correction sections 19 and 42, contour correction sections 20 and 43, and YCrCb-RGB conversion sections 21 and 44 to perform various processes. Although the processing is executed, the present invention is not limited to this, and a configuration in which these processings are executed by hardware may be adopted.
[0151]
In the case where the control units 8 and 27 are provided with the above-described units according to a predetermined program, for example, data is read from and written to an external recording medium such as a flexible disk, CD-R, or DVD-R. PC 26 is converted to RGB-YCrCb conversion units 9 and 32, block setting units 10 and 33, histogram generation units 11 and 34, block background level setting units 12 and 35, smoothing processing units 13 and 36, Background level calculation units 14 and 37, first gradation correction processing units 15 and 38, noise removal / connection units 16 and 39, first saturation correction units 17 and 40, second gradation correction processing units 18 and 41, 2 The program is installed from a recording medium on which a program for functioning as the saturation correction units 19 and 42, the outline correction units 20 and 43, and the YCrCb-RGB conversion units 21 and 44 is recorded. So as to Lumpur, PC 26 is by installing the program, may functionally configured with a like RGB-YCrCb conversion unit 9, 32.
[0152]
The main inventions disclosed in the present specification are summarized below.
[0153]
[Supplementary Note 1] The image is divided into a plurality of blocks including a plurality of pixels, a background level is set for each of the divided blocks, and a background level of a predetermined block in a block group including a plurality of blocks is set in the block group. A smoothing process is performed to set the average value of the background level in a predetermined number of blocks among the other blocks, and the block is located between the center pixel of each block and the center pixel using the background level calculated by the smoothing process. An image processing method comprising calculating a background level of a pixel.
[0154]
[Supplementary Note 2] The image processing method according to Supplementary Note 1, wherein the predetermined block is a block located farthest from the center of the image among blocks of a block group to which the predetermined block belongs. .
[0155]
[Supplementary Note 3] The image processing method according to Supplementary Note 1 or 2, wherein the smoothing process is performed in order from a block group located at the center of the image to a block group located on a peripheral side.
[0156]
[Supplementary Note 4] In the smoothing process, the background level of the predetermined block is set to an average value of the background levels of a plurality of blocks having lower luminance among the background levels of the blocks in the block group. 4. The image processing method according to any one of supplementary notes 1 to 3, wherein
[0157]
[Supplementary Note 5] Dividing means for dividing an image into a plurality of blocks including a plurality of pixels, setting means for setting a background level for each of the divided blocks, and a background level of a predetermined block in a block group including a plurality of blocks Using a background level calculated by the smoothing processing means to perform a smoothing process to make the average value of the background level in a predetermined number of blocks among the other blocks in the block group, each block An image processing apparatus comprising: a calculating unit configured to calculate a background level of a central pixel of the image and a pixel located between the central pixels.
[0158]
[Supplementary Note 6] The smoothing processing means sets the predetermined block so as to be located farthest from the center of the image among blocks of a block group to which the predetermined block belongs. The image processing device according to attachment 5.
[0159]
[Supplementary Note 7] The supplementary note 5 or 6, wherein the smoothing processing unit performs the smoothing process in order from a block group located at the center of the image to a block group located on a peripheral side. Image processing device.
[0160]
[Supplementary Note 8] The smoothing processing unit may set the background level of the predetermined block to an average value of the background levels of a plurality of blocks having lower luminance among the background levels of the blocks in the block group. The image processing apparatus according to any one of claims 5 to 7, wherein:
[0161]
[Supplementary Note 9] In the image processing apparatus,
Dividing means for dividing an image into a plurality of blocks each including a plurality of pixels;
Setting means for setting a background level for each divided block;
Smoothing processing means for performing a smoothing process for setting the ground level of a predetermined block in a block group including a plurality of blocks to an average value of ground levels in a predetermined number of blocks among other blocks in the block group;
A program for mounting a function as calculation means for calculating a background level of a center pixel of each block and pixels located between the center pixels using the background level calculated by the smoothing process.
[0162]
[Supplementary Note 10] In the image processing apparatus,
Dividing means for dividing an image into a plurality of blocks each including a plurality of pixels;
Setting means for setting a background level for each divided block;
Smoothing processing means for performing a smoothing process for setting the ground level of a predetermined block in a block group including a plurality of blocks to an average value of ground levels in a predetermined number of blocks among other blocks in the block group;
Using the background level calculated by the smoothing process, a program for mounting a function as a calculation unit for calculating a background level of a center pixel of each block and a pixel located between the center pixels is recorded. A recording medium readable by the image processing device.
[0163]
【The invention's effect】
According to the first or fifth aspect of the present invention, a background level is set for each of blocks obtained by dividing an image into a plurality of blocks including a plurality of pixels, and the background level of a predetermined block in a block group including a plurality of blocks is set. Is set to the average value of the background level in a predetermined number of blocks among the other blocks in the block group, so that the background level of the block having the pixel in which the reflected light or the like of the electric light is imaged is obtained. In addition, it is possible to reduce the difference from the background level of another block that does not capture the reflected light of the electric light or the like.
[0164]
Thereby, when calculating the background level of the center pixel of each block and the pixels located between the center pixels using the background level calculated by the smoothing process, the center pixel of each block and the space between the center pixels are calculated. It is possible to prevent the background level of the located pixel from being significantly different from the original background level, and as a result, the background removal processing can be appropriately performed.
[0165]
According to the invention described in claim 2, since the predetermined block in the block group is set to be located at a position farthest from the center of the image among the blocks of the block group to which the predetermined block belongs, The background level is calculated in a block closer to the center of the image where the reflected light of the electric light is less likely to be reflected. Thus, the background level of each pixel can be set more appropriately.
[0166]
According to the invention described in claim 3, the smoothing processing is performed in order from the block group located at the center of the image to the block group located on the peripheral side. The smoothing process is performed from the pixel located at the center of the image that is less likely to be captured. This makes it possible to more appropriately set the background level of the pixel.
[0167]
According to the invention described in claim 4, the smoothing process includes setting the background level of the predetermined block to the background level of a plurality of blocks having lower luminance among the background levels of the blocks in the block group. Since the average value is used, even if there is a block in which the reflected light of the electric light or the like is captured in the block group, the smoothing process using this block is avoided as much as possible. As a result, the background level of the pixel can be set more appropriately.
[Brief description of the drawings]
FIG. 1 is a control block diagram of a digital camera employing an image processing method according to the present invention.
2A is a diagram illustrating an example of an image, FIG. 2B is a diagram illustrating a state in which the image illustrated in FIG. 2A is divided into a plurality of blocks, and FIG. FIG.
FIG. 3 is a diagram illustrating examples of various histograms.
FIG. 4 is a diagram for explaining processing for setting a block base level.
FIG. 5 is a diagram illustrating a smoothing process.
FIG. 6 is a diagram illustrating a process of calculating a pixel base level.
FIG. 7 is a diagram illustrating a correction function of gradation correction performed for each pixel using a calculated pixel background level.
FIG. 8 is a diagram for explaining noise removal / connection processing by a noise removal / connection unit;
FIG. 9 is a diagram illustrating a correction function of gradation correction.
FIG. 10 is a flowchart illustrating a procedure of an image processing method according to the present invention.
11 is a flowchart illustrating a detailed procedure of a block base level setting process by a block base level setting unit in step # 5 of FIG. 10;
FIG. 12 is a flowchart illustrating a detailed procedure of a smoothing process by a smoothing processing unit in Step # 6 of FIG. 10;
FIG. 13 is a flowchart showing a detailed procedure of noise removal / connection processing by the noise removal / connection unit in step # 9 of FIG. 10;
FIG. 14 is a block diagram of a system configured to capture an image and store the image or output the image to recording paper or the like when the function of executing image processing according to the present invention is mounted on a personal computer. is there.
[Explanation of symbols]
8, 27 control unit 9, 32 RGB-YCrCb conversion unit 10, 33 block setting unit 11, 34 histogram generation unit 12, 35 block background level setting unit 13, 36 smoothing processing unit 14, 37 pixel background level calculation unit 15, 38 first tone correction processing section 16, 39 noise removal / connection section 17, 40 first saturation correction section 18, 41 second tone correction processing section 19, 42 second saturation correction section 20, 43 contour correction section 21,44 YCrCb-RGB converter

Claims (5)

Divide the image into multiple blocks of multiple pixels,
Set the background level for each divided block,
Performs a smoothing process to set the ground level of a predetermined block in a block group including a plurality of blocks to an average value of the ground level in a predetermined number of blocks among other blocks in the block group,
An image processing method, wherein a background level of a central pixel of each block and pixels located between the central pixels is calculated using the background level calculated by the smoothing process. The image processing method according to claim 1, wherein the predetermined block is a block located farthest from the center of the image among blocks of a block group to which the predetermined block belongs. 3. The image processing method according to claim 1, wherein the smoothing processing is performed in order from a block group located at a center of the image to a block group located on a peripheral side. 4. The smoothing process is characterized in that the background level of the predetermined block is an average value of the background levels of a plurality of blocks having lower luminance among the background levels of the blocks in the block group. The image processing method according to claim 1. Dividing means for dividing an image into a plurality of blocks each including a plurality of pixels;
Setting means for setting a background level for each divided block;
Smoothing processing means for performing a smoothing process for setting the ground level of a predetermined block in a block group including a plurality of blocks to an average value of ground levels in a predetermined number of blocks among other blocks in the block group;
An image processing apparatus comprising: a calculating unit configured to calculate a center pixel of each block and a background level of a pixel located between the center pixels using the background level calculated by the smoothing process.

Images