JP4038976B2 - Image processing apparatus, image processing method, and computer-readable recording medium storing image processing program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus, an image processing method, and a computer-readable recording medium on which an image processing program is recorded, and in particular, an image processing apparatus, an image processing method, and an image processing for removing noise caused by the influence of fading or fading The present invention relates to a computer-readable recording medium on which a program is recorded.
[0002]
[Prior art]
In recent years, with the spread of digital cameras, image scanners, and the like, there are increasing opportunities for taking images as digital data and storing or using them. Handling images as digital data is advantageous in that the images can be easily corrected or processed, and that predetermined images can be managed collectively using an electronic album or the like. For this reason, for example, reading from a film with a film scanner, or reading a developed photograph of a film with a digital camera or an image scanner, converting it into digital data, saving it, and managing it as an electronic album Is possible.
[0003]
However, when the film itself or a photograph developed from the film is exposed to sunlight or a fluorescent lamp, the film fades with time and the photograph fades. A method for correcting the faded or faded image is described in JP-A-11-136533. According to this method, a standard pattern for fading correction of each color of R (red), G (green), and B (blue) is preliminarily printed on a film, and the standard pattern burned when it is read by a film scanner or the like. Is determined, and a fading correction process is performed based on the determined fading level.
[0004]
Also, images taken under fluorescent or tungsten lamps are affected by color cast, and a technique for matching the dynamic range of each RGB component is known to eliminate the effect of this color cast. Yes.
[0005]
[Problems to be solved by the invention]
However, the fading correction method described in the above-mentioned Japanese Patent Application Laid-Open No. 10-136533 has a problem that a standard pattern for fading correction needs to be burned in advance on a film for taking an image, and has been developed from the film. A standard pattern for fading correction cannot be obtained from a photograph, and fading correction cannot be performed.
[0006]
Also, the method of removing color cast cannot correct an image that has faded due to deterioration of a developed photograph over time. This will be specifically described. FIG. 10 is a diagram illustrating an example of an RGB histogram of a color cast image taken under tungsten light. Referring to FIG. 10, the horizontal axis represents the luminance value, and the vertical axis represents the frequency of the pixel. Although the distribution shape of the RGB histogram is similar, the width and position of the distribution are different for each component of RGB. As described above, the color cast image has a feature that the distribution shape and the position of the distribution are different although the distribution shape of the histogram of each component is similar.
[0007]
FIG. 11 is a diagram illustrating an example of an RGB histogram of an image obtained by performing color cast correction on a color cast image taken under tungsten light. Referring to FIG. 11, correction is made so that the distribution shapes of the histograms of the RGB components are similar. In the correction of the color cast image, the maximum value and the minimum value of the histogram of each RGB component are obtained, and the density conversion is performed so that they match. As a result, the color cast image is corrected for the width and position of the histogram distribution of each component, and the color cast state is improved.
[0008]
FIG. 12 is a diagram illustrating an example of an RGB histogram of a fading image that has faded due to aging of the film. Referring to FIG. 12, in the faded image, not only the distribution width and position of the histogram of each RGB component but also the distribution shape of the histogram is different.
[0009]
FIG. 13 is a diagram illustrating an example of an RGB histogram of an image obtained by performing color cast correction processing on a fading image that has faded due to aging deterioration of the film. Referring to FIG. 13, the width and position of the distribution of the histogram of each RGB component match, but the shape of the distribution is different. For this reason, even if the color cast correction process is performed on the fading image, the influence of fading cannot be removed.
[0010]
The present invention has been made to solve the above-mentioned problems, and one of the objects of the present invention is an image processing apparatus capable of removing noise caused by the influence of fading or fading included in image data. And providing an image processing method.
[0011]
Another object of the present invention is a computer-readable recording medium on which an image processing program for causing a computer to execute image processing capable of removing noise caused by the influence of fading or fading included in image data is recorded. Is to provide.
[0012]
[Means for Solving the Problems]
  According to one aspect of the present invention, the above-mentioned object is achieved., PaintingImage data with multiple color components per elementAn image processing apparatus for performing fading correction processing on an image dataReceiving means for receiving received image data, distribution obtaining means for obtaining a luminance value distribution for each of a plurality of color components, and means for obtaining a maximum value and a minimum value of a luminance value for each of the plurality of color components; For each of multiple color componentsN equal value acquisition means for obtaining an n equal value obtained by dividing the maximum value and the minimum value into n equal values by the number of pixels (n is an integer of 3 or more).And the maximum and minimum luminance values obtained for each of the plurality of color components, andThe n equal value matches the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.Correction means for correcting the luminance value distribution of a plurality of color components.
  Preferably, the correction unit obtains a difference between the maximum value and the minimum value obtained for each of the plurality of color components, uses the component having the maximum difference as a reference component, and sets the maximum and minimum values of the reference component, n, and the like. Interpolation is performed so that the maximum value, the minimum value, and the n equal value of other components are matched with the minute value.
[0013]
  According to the present invention, for the received image data, a luminance value distribution is obtained for each of the plurality of color components, and for each of the plurality of color components, a maximum value and a minimum value of the luminance value are obtained, and each of the plurality of color components is obtained. about,The n equal value obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more).Desired. The maximum value, the minimum value, and the luminance value obtained for each of the plurality of color componentsThe n equal value matches the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.The luminance value distribution of the plurality of color components is corrected. Therefore, the luminance value distribution can be matched between the color components. As a result, it is possible to provide an image processing apparatus capable of removing noise caused by the influence of fading or fading included in the image data by matching the luminance value distributions of the color components of the image data.
[0018]
  Preferably, the image processing apparatus converts a received image into a color signal on a uniform color space, and based on the converted color signal.By detecting that the hue histogram is biased to a specific hue, or that the saturation histogram is biased to low-saturation partsAnd determining means for determining whether the received image data is image data obtained by capturing a faded or faded image, and the correction means captures an image where the received image data is faded or faded Determined as image data obtainedDepending onThe luminance value distribution of a plurality of color components is corrected.
[0019]
  According to the present invention, the received image is converted into a color signal in the uniform color space, and the converted color signal is used.By detecting that the hue histogram is biased to a specific hue or that the saturation histogram is biased to a low-saturation partIt is determined whether the received image data is image data obtained by capturing a faded or faded image, and the received image data is determined to be image data obtained by capturing a faded or faded image WasDepending onThe luminance value distribution of the plurality of color components is corrected. For this reason, it is possible to appropriately correct only the faded image.
[0020]
  Preferably, the image processing apparatus further includes an input unit for receiving whether the received image data is image data obtained by capturing a faded or faded image, and the correction unit is configured to receive the received image. The fact that the data is the image data obtained by capturing the faded image was receivedDepending onThe luminance value distribution of a plurality of color components is corrected.
[0021]
  According to the present invention, it is received whether the received image data is image data obtained by capturing a faded or faded image, and the image obtained by capturing the faded image of the received image data Received dataDepending onThe luminance value distribution of the plurality of color components is corrected. For this reason, it is possible to appropriately correct only the faded image.
[0022]
  According to another aspect of the invention, PaintingImage data with multiple color components per elementAn image processing method for performing fading correction processing on an image dataA step of obtaining a luminance value distribution for each of a plurality of color components for the received image data, a step of obtaining a maximum value and a minimum value of luminance values for each of the plurality of color components, and a plurality of colors For each ingredient,N equal values obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more).Determining a maximum value, a minimum value, and a luminance value obtained for each of the plurality of color components;The n equal value is matched with the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.Correcting the luminance value distribution of the plurality of color components.
[0023]
  According to the present invention, for the received image data, a luminance value distribution is obtained for each of the plurality of color components, and for each of the plurality of color components, a maximum value and a minimum value of the luminance value are obtained, and each of the plurality of color components is obtained. about,The n equal value obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more).Desired. The maximum value, the minimum value, and the luminance value obtained for each of the plurality of color componentsThe n equal value matches the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.The luminance value distribution of the plurality of color components is corrected. Therefore, the luminance value distribution can be matched between the color components. As a result, it is possible to provide an image processing method capable of removing noise caused by the influence of fading or fading included in image data by matching the luminance value distributions of the color components of the image data.
[0024]
  According to yet another aspect of the invention, PaintingImage data with multiple color components per elementA computer-readable recording medium storing an image processing program for causing a computer to execute the fading correction processing ofA step of obtaining a luminance value distribution for each of a plurality of color components for the received image data, a step of obtaining a maximum value and a minimum value of luminance values for each of the plurality of color components, and a plurality of colors For each ingredient,N equal values obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more).Determining a maximum value, a minimum value, and a luminance value obtained for each of the plurality of color components;The n equal value matches the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.An image processing program for causing a computer to execute a step of correcting luminance value distributions of a plurality of color components is recorded.
[0025]
  According to the present invention, for the received image data, a luminance value distribution is obtained for each of a plurality of color components, and a maximum value and a minimum value of the luminance value are obtained for each of the plurality of color components, and each of the plurality of color components is obtained. about,The n equal value obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more).Desired. Then, the maximum value, the minimum value, and the brightness value obtained for each of the plurality of color componentsThe n equal value matches the maximum value, the minimum value, and the n equal value of the reference component determined from the plurality of color components based on the maximum value and the minimum value, respectively.The luminance value distribution of the plurality of color components is corrected. Therefore, the luminance value distribution can be matched between the color components. As a result, image processing for causing a computer to execute image processing that can eliminate noise caused by fading or fading included in image data by matching the luminance value distribution of the color components of the image data with each other A computer-readable recording medium in which the program is recorded can be provided.
[0026]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the drawings, the same reference numerals indicate the same or corresponding members, and description thereof will not be repeated.
[0027]
FIG. 1 is a diagram illustrating a connection relationship between an image processing apparatus and an external apparatus connected to the image processing apparatus according to an embodiment of the present invention. Referring to FIG. 1, image processing apparatus 100 is configured with a personal computer. An image input device 150 typified by an image scanner and an image output device 160 for outputting an image from the image processing device 100 are connected to the image processing device 100.
[0028]
The image input device 150 may be a digital camera, a digital video camera, or the like. An image input from the image input device 150 to the image processing device 100 may be a still image or a moving image. In the case of a moving image, the image processing apparatus 100 performs image processing similar to that performed on a still image for each frame.
[0029]
The image output device 160 is a laser printer, an inkjet printer, or the like, and can output a full color image.
[0030]
The image processing apparatus 100 includes an external storage device, and can read an image processing program to be executed by the image processing apparatus 100 from a recording medium 170 such as a CD-ROM, a magneto-optical disk, a digital video disk, or a floppy disk. In this case, an image processing program for causing the image processing apparatus 100 to execute a fading correction process described later is recorded on the recording medium 170. Then, the image processing apparatus 100 can execute the program by reading the image processing program with the external storage device. Further, the color fading correction process described below may be performed by the image input device 150.
[0031]
FIG. 2 is a functional block diagram showing an outline of functions of the image processing apparatus according to the present embodiment. Referring to FIG. 2, image data composed of R (red), G (green), and B (blue) components is sent from image input device 150 to image processing device 100.
[0032]
The image processing apparatus 100 includes a scene determination processing unit 110 and an image correction processing unit 120. The image data received from the image input device 150 is sent to the scene determination processing unit 110 and the image correction processing unit 120.
[0033]
The scene determination processing unit 110 includes a color space conversion processing unit 111 for converting received image data into a uniform color space, and each of three components of hue, saturation, and lightness for the image data converted into the uniform color space. And a scene information determination unit 115 that determines scene information indicating the nature of the image data based on the three generated histograms.
[0034]
The color space conversion processing unit 111 converts image data represented by three components of RGB into a uniform color space represented by three components of hue, saturation, and lightness. The uniform color space in the Munsell color system is used for the uniform color space represented by hue, saturation, and lightness. Although the Munsell color system is used here, a display system such as L * a * b * color system, L * u * v * color system, CIE color system, DIN system, Ostwald display, etc. May be used.
[0035]
The histogram calculation unit 113 creates a histogram for each of hue, saturation, and lightness. The scene information determination unit 115 determines whether the image is a fading or fading image based on the histogram created by the histogram calculation unit 113.
[0036]
Note that there is no need to distinguish between a faded image obtained by imaging a faded film and a faded image obtained by imaging a faded photograph. Therefore, these are hereinafter collectively referred to as a faded image unless otherwise specified.
[0037]
As a method for determining scene information of image data, a histogram distribution of hue, saturation, and brightness is used. A fading image can be determined by detecting that the hue histogram is biased to a specific hue or that the saturation histogram is biased to a portion with low saturation. The bias of the histogram distribution is determined using the average value, median value, variance, maximum value, and minimum value.
[0038]
Further, since the fading image has a feature that there are few achromatic portions, there are very few achromatic (white, black, gray) regions in the region excluding the high lightness and low lightness of the lightness histogram. For this reason, whether or not the achromatic region is equal to or greater than a certain value may be added to the determination criterion.
[0039]
The image correction processing unit 120 performs backward correction color processing on the input image data. The image correction processing unit 120 generates a histogram for each of hue, saturation, and lightness of input image data, and calculates a maximum value, a minimum value, and an n-divided value in each of the generated histograms. A reference pixel value calculation unit 123 that converts the density value of each histogram.
[0040]
Next, the principle of the fading correction process will be described. The fading correction process is a process for matching the shape and range of the distribution in the histogram of the three RGB components. First, the maximum value and the minimum value of each component of RGB are obtained. Then, an n equal value obtained by dividing the histogram of each component into n equal parts (an integer of n ≧ 2) by the number of pixels is obtained. (n-1) pieces of n equal values are obtained. In addition, the n equal values have the same cumulative value between the RGB components by accumulating the number of pixels in order of increasing or decreasing luminance values. Then, n equal values are associated between RGB components.
[0041]
In the present embodiment, the case where n = 3 will be described for the sake of explanation. This n equal division is not limited to three equal parts, and is not limited to this as long as it is equal to or more than two equal parts. By increasing n, it is possible to improve the accuracy of the fading correction process. When the number of pixels is divided into three equal parts, there are two three equal values.
[0042]
Here, the maximum value of the R component is Rmax, the minimum value is Rmin, the maximum value of the G component is Gmax, the minimum value is Gmin, the maximum value of the B component is Bmax, and the minimum value is Bmin. Then, the R component trisection value is set to Rn1 and Rn2 in order from the lowest luminance value, the G component trisection value is set to Gn1 and Gn2 from the lower luminance value, and the B component trisection value is set. Bn1 and Bn2 are set in order from the smallest luminance value.
[0043]
At this time, when the pixels are arranged in ascending order of the pixel values in each of the R component, the G component, and the B component, the number of pixels existing between Rmin and Rn1 is the number of pixels existing between Gmin and Gn1, Or it becomes the same as the number of pixels existing between Bmin and Bn1. Similarly, the number of pixels existing between Rn1 and Rn2 is the same as the number of pixels existing between Gn1 and Gn2, or the number of pixels existing between Bn1 and Bn2. Further, the number of pixels existing between Rn2 and Rmax is the same as the number of pixels existing between Gn2 and Gmax or the number of pixels existing between Bn2 and Bmax.
[0044]
The density conversion process is performed on the histogram of each component so that the luminance values of the maximum value, the minimum value, and the n equal value obtained in this way are matched between the histograms of the three RGB components. Linear interpolation is used for this density conversion processing. At this time, the difference between the maximum value and the minimum value of each component is obtained, the component having the maximum difference is set as the reference component, and the maximum value of the other component is set to the maximum value, the minimum value, and the n equal value of the reference component. Then, the density conversion process is performed so that the minimum value and the n equal value match.
[0045]
FIG. 3 is a diagram illustrating an RGB histogram of a fading image before and after performing fading correction processing. FIG. 3A is a diagram showing an RGB histogram of a fading image, and FIG. 3B is a diagram showing an RGB histogram of an image after fading correction processing.
[0046]
Referring to FIG. 3A, among RGB components, the maximum value (Rmax, Gmax, Bmax), the minimum value (Rmin, Gmin, Bmin) and the trisection value (Rn1, Gn1, Bn1 and Rn2, The brightness values of Gn2 and Bn2) do not match. Further, the intervals of the maximum value, the minimum value, and the trisection value do not match among the RGB components. For example, the interval between Rn1 and Rn2, the interval between Gn1 and Gn2, and the interval between Bn1 and Bn2 are all different.
[0047]
Referring to FIG. 3B, with the R component as a reference component, the maximum value (Rmax) of the R component matches the maximum value (Gmax) of the G component and the maximum value (Bmax) of the B component. The G component minimum value (Gmin) and the B component minimum value (Bmin) coincide with the minimum value (Rmax), and the R component trisection value (Rn1 and Rn2) matches the G component trisection value (Gn1). , Gn2) and the B component's trisection values (Bn1 and Bn) are respectively matched so that the luminance values of the G component and B component are converted. For the pixels other than these pixels, linear interpolation based on the maximum value, the minimum value, and the trisection value is performed. The shape, range, and position of the histogram distribution for each of RGB are similar.
[0048]
FIG. 4 is a diagram for explaining linear interpolation used for density conversion processing. Referring to FIG. 4, the horizontal axis indicates the B component, the vertical axis indicates the R component, and the minimum value (Bmin, Rmin), the maximum value (Bmax, Rmax), and the trisection value ( Bn1, Bn2, Rn1, Rn2). Here, the reference component is described as an R component. B component minimum value (Bmin), R component minimum value (Rmin), B component maximum value (Bmax), R component maximum value (Rmax), B component trisection (Bn1, Bn2) By plotting points determined by the R component trisection value (Rn1, Rn2) and connecting each point with a straight line, a conversion straight line used for density conversion processing is obtained.
[0049]
The density conversion process does not convert the density with a straight line connecting the minimum value, the n-divided value, and the maximum value, but a point Amin determined by the minimum value (Bmin, Rmin) of two components and a tri-section value. A density conversion curve (tone curve) passing through the point An1 determined by Bn1 and Rn1, the point An2 determined by the trisection values Bn2 and Rn2, and the point Amax determined by the maximum values (Bmax, Rmax) may be used. . The tone curve is obtained by performing spline interpolation. Alternatively, it may be obtained using a high-dimensional function.
[0050]
FIG. 5 is a diagram showing a density conversion curve used for density conversion processing. Referring to FIG. 5, a spline-interpolated curve connecting the minimum value, n-equal value, and maximum value is represented. As described above, by performing density conversion using the spline-interpolated curve, smoothness of gradation is not lost, and appropriate interpolation processing can be performed. For this reason, it is possible to prevent the occurrence of a tone jump in which the gradation changes rapidly in an image subjected to fading correction.
[0051]
In the present embodiment, the description has been made assuming that n = 3. When n = 2, density conversion processing can be performed effectively when the shape of the histogram of each component is unimodal. However, if any one of the component histograms is not unimodal, for example, if it is bimodal, the bisection value (median value) and the two peak positions of the histogram will be greatly shifted. For this reason, the shape of the histogram cannot be made similar. In this embodiment, since n = 3, it is possible to cope with the case where the shape of the histogram is bimodal.
[0052]
FIG. 6 is a diagram for explaining a defect in the fading correction process when n = 2. 6A shows two histograms before density conversion processing, and FIG. 6B shows a histogram of an image subjected to density conversion processing with n = 2. For the sake of explanation, only the histograms of the R component and the B component are shown.
[0053]
As is apparent from FIG. 6A, the shape of the R component histogram is bimodal, and the shape of the B component histogram is unimodal. The position of the pixel Rn obtained by dividing the R component pixel value into two equal parts does not match the peak positions (R1, R2) of the two peaks in the R component histogram. Since the histogram of the B component is unimodal, the pixel position Bn divided into two equal parts and the peak position B1 of the histogram substantially coincide.
[0054]
Referring to FIG. 6B, the shapes of the R component histogram and the B component histogram are completely different, and the shapes of the histograms cannot be matched by the fading correction processing that bisects the image. As a result, the fading correction process is not effective.
[0055]
FIG. 7 is a diagram showing an RGB histogram of an image obtained by performing a fading correction process on the fading image of the RGB histogram shown in FIG. Referring to FIG. 7, the shapes, ranges, and positions of the histograms of the R, G, and B components are similar.
[0056]
FIG. 8 is a flowchart showing the flow of fading correction processing performed by the image processing apparatus according to the present embodiment. Referring to FIG. 8, in the fading correction process, first, RGB image data is input by image input device 150 (step S01). Then, the input RGB image data is converted into image data composed of hue, saturation and brightness. In this case, RGB image data is converted into a color signal in a color space defined by the Munsell color system.
[0057]
Then, the histogram calculation unit 113 creates histograms for hue, saturation, and brightness from the converted image data (step S03). Thereafter, the scene information determination unit 115 determines the scene information of the image (step S04). The scene information determination is a process for determining whether an image corresponds to a faded image. In many cases, a fading image is biased to a specific hue as a whole, and the color is generally thin and the saturation is low. Accordingly, the scene information determination unit 115 checks the degree of color bias from the hue histogram, and if it is determined that the color is biased to a specific hue, the scene information determination unit 115 biases to a low saturation region from the saturation histogram. It is checked whether or not When it is determined that the saturation histogram is biased to a low region as a whole, it is determined as a fading image.
[0058]
In addition, the fading image has a feature that there are few achromatic portions. Therefore, it may be determined whether or not the pixels included in the region excluding the high lightness region and the low lightness region have an achromatic color (white, black, gray) or more in a predetermined ratio. The determination method using the above-described hue histogram and saturation histogram and the method using the brightness histogram and hue histogram may be used alone, or may be determined in combination.
[0059]
In the present embodiment, the scene of the image is determined using the hue, saturation, and lightness histograms. However, it is also possible to determine using the histogram of each RGB component. In this case, the scene determination processing unit 110 is not necessary. When a scene is determined using a histogram of each component of RGB, for example, the histogram of each component of RGB of a fading image often differs greatly in distribution shape. By determining the difference in distribution, a fading image can be determined.
[0060]
More specifically, after performing linear interpolation that matches the maximum value and the minimum value of each of the RGB histograms, it is checked whether the histogram distribution shapes after the interpolation are similar. The criteria for determining whether or not they are similar may be to determine the similarity between the average value, variance value, and luminance value at which the frequency peaks. When it is determined that the distributions of the three histograms are not similar, it is determined as a fading image.
[0061]
As a result of the determination in step S04, it is determined whether the image is a fading image (step S05). If it is determined that the image is a fading image, the process proceeds to step S06, and if not, the process proceeds to step S09.
[0062]
In step S <b> 06, a histogram for each of the RGB components is created by the histogram calculation unit 121. Then, the reference pixel value calculation unit 123 calculates the maximum value, the minimum value, and the n (n = 3) equal value of the histogram of each component created (step S07). The n equal value is a luminance value of pixels in which the cumulative number of pixels from the maximum value is the same when the pixels are arranged in descending order of luminance value in each histogram. Or it is a luminance value of a pixel in which the cumulative number of pixels from the minimum value is the same when the pixels are arranged in ascending order of luminance value.
[0063]
For example, when the number of pixels is 300, the number of pixels divided into three equal parts is 100. Therefore, the luminance value of the pixel with the 100th largest luminance value and the luminance value of the pixel with the 200th largest luminance value from the maximum value of each histogram are obtained from the histogram of each component.
[0064]
Then, density conversion processing is performed on each component so that the obtained minimum value, maximum value, and n-equal value match between the histograms of the three components (step S08). In this density conversion process, the difference between the maximum value and the minimum value of the histogram of each component is obtained, the component having the maximum difference is set as the reference component, the reference component maximum value, the minimum value, and the n equal value. Linear interpolation is performed so that the minimum value, the maximum value, and the n equal value of the components are matched.
[0065]
In the next step S09, the density-converted image data is output to the image output device 160.
[0066]
In the present embodiment, the scene of the image data input in step S02 to step S04 is automatically determined, but an instruction signal indicating that the image is a faded image is input from the keyboard by the user of the image processing apparatus 100. You may make it do.
[0067]
FIG. 9 is a diagram illustrating an example of a graphic user interface (GUI) for designating a fading image displayed on the image processing apparatus according to the present embodiment. Referring to FIG. 9, scene information such as fading, backlighting, underexposure, sunset, and people is exemplified in the scene selection field. The scene information is input by instructing a check column provided corresponding to the exemplified scene information using a pointing device such as a mouse or a keyboard. In the present embodiment, an instruction signal indicating that the image is a faded image is input as the scene information by instructing a check column provided corresponding to the fade.
[0068]
The image before the fading correction process is displayed in the “pre-correction image” area, and the image after the fading correction process is arranged and output in the “post-correction image” area.
[0069]
As described above, the image processing apparatus according to the present embodiment converts the density so as to match the shape, range, and position of the luminance value histogram of the RGB components of the image data. For this reason, it is possible to easily remove noise generated by the influence of fading or fading included in the image data.
[0070]
In the density conversion process, the density conversion process is performed so that the luminance values of the RGB component maximum value, the minimum value, and the n-equal value are matched, so that the luminance value distribution of a plurality of components has a complicated shape. In addition, the luminance value distributions of a plurality of components can be easily matched.
[0071]
Furthermore, since the scene determination processing unit 110 automatically determines whether or not the image is a faded image using the image data converted into hue, saturation, and brightness, the fading correction process is performed only on the faded image. be able to.
[0072]
The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.
[Brief description of the drawings]
FIG. 1 is a diagram showing a connection relationship between an image processing apparatus and an external apparatus connected to the image processing apparatus in one embodiment of the present invention.
FIG. 2 is a functional block diagram showing an outline of functions of an image processing apparatus according to the present embodiment.
FIG. 3 is a diagram illustrating an RGB histogram of a fading image before and after performing fading correction processing.
FIG. 4 is a diagram for explaining linear interpolation used for density conversion processing;
FIG. 5 is a diagram showing a density conversion curve used for density conversion processing.
FIG. 6 is a diagram for explaining a defect in fading correction processing when n = 2.
7 is a diagram showing an RGB histogram of an image obtained by performing a fading correction process on the fading image of the RGB histogram shown in FIG. 12; FIG.
FIG. 8 is a flowchart showing a flow of fading correction processing performed by the image processing apparatus according to the present embodiment.
FIG. 9 is a diagram illustrating an example of a graphic user interface for designating a fading image displayed on the image processing apparatus according to the present embodiment.
FIG. 10 is a diagram showing an example of an RGB histogram of a color cast image taken under tungsten light.
FIG. 11 is a diagram illustrating an example of an RGB histogram of an image obtained by performing color cast correction on a color cast image photographed under tungsten light.
FIG. 12 is a diagram illustrating an example of an RGB histogram of a fading image that has faded due to aging of a film.
FIG. 13 is a diagram illustrating an example of an RGB histogram of an image obtained by performing color cast correction processing on a fading image that has faded due to aging deterioration of a film.
[Explanation of symbols]
100 image processing apparatus, 110 scene determination processing unit, 111 color space conversion processing unit, 113, 121 histogram calculation unit, 115 scene information determination unit, 120 image correction processing unit, 123 reference pixel value calculation unit, 125 density conversion unit, 150 Image input device, 160 Image output device, 170 Recording medium.

Claims (6)

An image processing apparatus for performing fading correction processing on image data having a plurality of color components for each pixel ,
Receiving means for receiving the image data ;
Distribution acquisition means for obtaining a luminance value distribution for each of the plurality of color components for the received image data;
Means for obtaining a maximum value and a minimum value of luminance values for each of the plurality of color components;
For each of the plurality of color components, n equal value acquisition means for obtaining an n equal value obtained by dividing the maximum value and the minimum value by n pixels (n is an integer of 3 or more) ;
The maximum value obtained for each of the plurality of color components, the minimum value and the n equal value, said maximum reference component defined from among the plurality of color components based on the maximum value and the minimum value An image processing apparatus comprising: a correction unit that corrects the luminance value distribution of the plurality of color components so as to match the value, the minimum value, and the n equal value . The correction unit obtains a difference between the maximum value and the minimum value obtained for each of the plurality of color components, sets a component having the maximum difference as the reference component, and sets the maximum value of the reference component as The image processing apparatus according to claim 1, wherein interpolation is performed so that the minimum value and the n equal value match the maximum value, the minimum value, and the n equal value of another component. Conversion means for converting the received image into a color signal on a uniform color space;
Being biased to said converted color signal based rather hue histogram specific hue, or by detecting that the saturation histogram is biased to the lower part of the saturation, the received image data is fading Or a determination means for determining whether the image data is obtained by capturing a faded image;
Wherein the correction means, the received image data in response especially is determined that the image data obtained by imaging the fading or faded images, corrects the luminance value distribution of the plurality of color components, according to claim The image processing apparatus according to 1. An input means for receiving whether the received image data is image data obtained by capturing a faded or faded image;
The correction unit corrects the luminance value distribution of the plurality of color components in response to receiving that the received image data is image data obtained by capturing a faded or faded image. The image processing apparatus according to claim 1. An image processing method for performing a fading correction process on image data having a plurality of color components for each pixel ,
Receiving the image data ;
Obtaining a luminance value distribution for each of the plurality of color components for the received image data;
Obtaining a maximum value and a minimum value of luminance values for each of the plurality of color components;
For each of the plurality of color components, obtaining an n equal value obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more) ;
The maximum value obtained for each of the plurality of color components, the minimum value and the n equal value, said maximum reference component defined from among the plurality of color components based on the maximum value and the minimum value And correcting the luminance value distribution of the plurality of color components so as to match the value, the minimum value, and the n-divided value, respectively . A computer-readable recording medium recording an image processing program for causing a computer to execute a fading correction process of image data having a plurality of color components for each pixel ,
Receiving the image data ;
Obtaining a luminance value distribution for each of the plurality of color components for the received image data;
Obtaining a maximum value and a minimum value of luminance values for each of the plurality of color components;
For each of the plurality of color components, obtaining an n equal value obtained by dividing the maximum value and the minimum value into n equal parts by the number of pixels (n is an integer of 3 or more) ;
The maximum value obtained for each of the plurality of color components, the minimum value and the n equal value, said maximum reference component defined from among the plurality of color components based on the maximum value and the minimum value value, the minimum value and the n equally divided values of the computer readable and correcting the luminance value distribution of the plurality of color components recorded an image processing program to be executed by the computer to match each record Medium.

Images