JP4379422B2 - Image processing device - Google Patents

Description

  The present invention relates to an image correction process for a color image.

In an image processing apparatus, it is generally preferable to correct an input color image. Therefore, various corrections such as color balance adjustment and color correction are performed. For example, color cast refers to a situation where a specific color covers the entire image and the color balance is lost. Specifically, a typical example can be seen in a photograph taken without using a strobe under a fluorescent lamp. However, it can actually occur in various other situations.
JP-A-9-149271

  In image processing, as a technique for correcting the color balance normally, it is well known that the brightest part is pure white and the darkest part is black. This method is simple and effective. However, depending on the image, there may be many exceptions such as an adverse effect or no effect.

  An object of the present invention is to provide an image processing apparatus capable of adjusting the color balance of an image more effectively.

An image processing apparatus according to the present invention includes a detection unit that detects the presence or absence of color fogging in color image data to be processed, and a fog color for each pixel that constitutes color image data detected as having color fogging by the detection unit. and certification means for certifying, among the color components consisting RGB of the pixel, and a subtraction means for performing subtraction processing to reduce the value of the color component corresponding to the recognized as a fog color by the certification unit color The Here, the recognition means recognizes the color corresponding to the color component having the highest value among the color components composed of RGB of the pixel as a fog color, and the subtraction means regards the color component composed of RGB of the pixel. Of these, the value of the color component corresponding to the fog color certified by the certification means is subtracted so as to be the same value as the value of the color component having the next largest value.

  Preferably, a contrast correction unit that corrects the contrast of the color image data after the subtraction processing by the subtraction unit is further provided.

The computer-readable recording medium according to the present invention includes a step of detecting the presence or absence of color fog for color image data to be processed, and a fog color for each pixel constituting the color image data detected as having color fog. Storing a program for causing the computer to execute a step of reducing a value of a color component corresponding to a color recognized as a fog color among the color components composed of RGB of the pixel. The computer-readable recording medium . Here, in the step of recognizing the cover color, the program recognizes a color corresponding to the color component having the highest value among the color components composed of RGB of the pixel as a cover color, and performs the subtraction process. In the step, the value of the color component corresponding to the certified fog color among the color components composed of RGB of the pixel is subtracted so as to be the same value as the value of the color component having the next largest value. , Certifying the cover color and performing the subtraction process.

Preferably, the program further causes the computer to execute a step of correcting the contrast of the color image data after the subtraction process in the step of performing the subtraction process.

  According to the present invention, since the subtraction process for reducing the image data of the color recognized as the color cast is performed, the color balance of the image can be adjusted more effectively.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings, the same reference symbols denote the same or equivalent.

  An information processing apparatus (hereinafter referred to as a system) according to an embodiment of the present invention is configured around a control apparatus 1 (computer) that includes a central processing unit (hereinafter referred to as a CPU) and controls the entire system. In FIG. 1, arrows indicate the flow of data. The display 2 displays images or characters and displays various screens for operation. The keyboard 3 and mouse 4 perform various inputs and instruction operations. The flexible disk 5 and the hard disk 6 are data storage media for storing and storing image data and attribute information thereof, and the system includes a flexible disk drive and a hard disk drive that access the flexible disk 5 and the hard disk 6, respectively. The printer 7 prints image data, a drawing created by image editing, and the like on a sheet. The scanner 8 reads image data from a document. The CD-ROM 9 is a recording medium that stores a large amount of various data, and the system includes a CD-ROM drive that accesses the CD-ROM 9. Further, a speaker 10 for audio output and a microphone 11 for audio input are connected. An image processing program to be described later is read from an external recording medium such as a CD-ROM.

  FIG. 2 is a block diagram centering on the control device 1. The control device 1 is configured with the CPU 201 as the center, and via a data bus 220 connected to the CPU 201, a ROM 203 that stores various processing programs, a RAM 204 that stores various data and programs, and displays images or characters. Display control circuit 205 for display 2, keyboard control circuit 206 for transfer control of input from keyboard, mouse control circuit 207 for transfer control of input from mouse 4, flexible disk drive control circuit 208 for controlling flexible disk drive, hard disk A hard disk control circuit 209 that controls the drive, a printer control circuit 210 that controls output to the printer 7, a scanner control circuit 211 that controls the scanner 8, and a CD-ROM drive system that controls the CD-ROM drive. Circuit 212, a speaker control unit 213 controls the speaker 10 and, a microphone control circuit 214 for controlling the microphone 11 are respectively connected.

  The clock circuit 202 supplies various clock signals necessary for operating this system. Further, an expansion slot 215 for connecting various expansion boards via the data bus is connected. Note that an SCSII board may be connected to the expansion slot 215 to connect a flexible disk drive, a hard disk drive, a scanner, a CD-ROM drive, or the like.

  In this system, the flexible disk 5 and the hard disk 6 are used as recording media, but other information recording media such as MO may be used. Further, although the scanner 8 and the CD-ROM 9 are used as image data input devices, other input devices such as a still video camera may be used. Further, although the printer 7 is used as an output device, other output devices such as a digital copier may be used.

  This system performs automatic editing processing of full-color images. Here, in the image correction process, the image data stored in the image memory is corrected. When a correction processing start command is input by the user, an image correction processing routine is activated to perform image correction processing. Image correction includes image file input / output processing, image display processing, graphic user interface (GUI) processing for correction operation, error processing in GUI processing, resolution change processing, and the like. To do.

  In the automatic image correction processing in this system, image correction is performed on the image data. The image correction routine includes color balance processing that is the subject of the present invention. When a correction processing start command is input by the user, this routine is activated to automatically perform image correction processing. The automatic image correction includes an image file input / output process, an image display process, a graphic user interface (GUI) process for correction operation, an error process in the GUI process, and a resolution change process. To do.

  FIG. 3 shows a main flow of image correction by the control device 1. In this image correction, correction is performed so as to maximize the overlap of RGB histograms when correcting the color balance. Japanese Patent Application Laid-Open No. 9-149271 describes image processing for performing simple expansion using a histogram. However, this does not focus on how the histograms overlap.

  First, an image correction target file is selected (step S10, hereinafter “step” is omitted), and an output destination of the file after image correction is selected (S12).

  Next, the selected files are read sequentially to perform image correction. First, a variable n representing an image file is set to 1 (S14), and the nth image file is input (S16). Next, image correction is executed (S18) and output to the nth image file (S20). In the present embodiment, as will be described later, in the image correction execution (S18), the image is corrected so that the overlap of RGB histograms is maximized.

  Next, the variable n is incremented (S22), and it is determined whether or not the processing of all input files is completed (S24). If not, the process returns to S16 and the above process is repeated for all the selected image files.

  FIG. 4 shows a flow of execution of image correction (S18 in FIG. 3). Here, the image is corrected so that the histogram overlap is maximized. First, a histogram of each RGB plane is created for input image data (S30), and a peak value of the RGB histogram is calculated (S32). Next, translation is performed to point O of the RGB histogram (S34). Next, the histograms are multiplied so that the peak values of the RGB histograms coincide (S36). Next, the overlapping area of the RGB histograms is calculated (S38). Next, the parallel movement amount is calculated so that the overlapping area of the RGB histograms is maximized (S40). This parallel movement amount calculation was performed as follows. First, the overlapping area of the RGB histograms is calculated. Next, the center of gravity of RGB is calculated, and movement is performed so that the other two of the center of gravity of RGB have the largest value. Next, the other two moved are moved one step at a time to calculate the overlap area, and if smaller, return. Next, color balance correction according to the curve of the RGB histogram is performed on each of the RGB planes (S42). This correction is performed as follows. Compare the original histogram and the histogram with the largest overlap area, calculate the parallel movement amount and magnification for each gradation for each RGB plane, and perform processing according to the gradation for each pixel .

Next, a second embodiment of the present invention will be described. In the present embodiment, in the execution of image correction (FIG. 3, S18), color cast is detected and correction is performed .

  FIG. 5 shows a flow of execution of image correction in the second embodiment. First, a color cast is detected (S50). Color cast is a situation where a specific color covers the entire image and the color balance is lost.

  This color fog detection was performed as shown in FIG. First, the image is converted from RGB data to HSL data (S500). Next, a portion having an area color having L of about 15 to 85 is extracted from the image (S502). In determining color cast, the exclusion of specific color objects must be taken into account (exclusion of certain colors). Therefore, it is next determined whether or not 85% or more of all the pixels are gathered in a specific region (width of about 20 degrees) in the hue (H) (S504). If they are gathered in this area, it is determined that the color is not a color cast but a constant color. Next, it is determined whether or not 80% or more of all the pixels are gathered in a specific region (width of about 50 degrees) in the hue (H) (S506). If it is gathered in this area, it is determined that the color cast is that color. Further, another determination is made regarding the green fog (S508). Since the green fog tends to spread more than the other color fog in the hue (H), the area width is increased to about 90 degrees and re-determination is performed. If it falls within this range, it is determined that the head is green. Finally, the color determined as the color cast is converted into the RGB space, and the highest color among the R, G, and B values is recognized as the cast color (S510).

Referring back to FIG. 5, then performs a subtraction of the pressurized first time color (S52). Specifically, each pixel has an RGB value, but the highest color among the RGB values is recognized as a cover color. Here, assuming that RGB = (200, 180, 130), in this case, the fogging color is R and the value thereof is 200. Subtraction here means subtracting 20 from 200 to 180. RGB = (180, 180, 130).

  The movement of the histogram in this process will be specifically described. To simplify the story, consider only RG among RGB. FIG. 7 shows an example of the original histogram. In FIG. 7, the R histogram extends to a high place. Now consider a point P having a value of RG = (230,180). If P is present in the region represented by point A in FIG. 7 with respect to R, reducing R to the same level as G reduces the number of points corresponding to A on the histogram, resulting in region D. It will move to somewhere inside.

Then, by inverting the RGB to create a data of YMC, subtracting the here or Laka the first time in color (S54). It is possible to correct too many colors of RGB by the processing up to step S52 described above. However, since too few colors cannot be corrected, the correction is realized by reversing RGB to create a YMC and performing the same processing.

  Next, the maximum value and the minimum value of each RGB plane are obtained, the respective enlargement ratios are calculated (S56), and the enlargement is performed according to the enlargement ratios of the respective RGB planes (S58). In this image correction, since the color cast is removed in the direction of removing the color, the whole image tends to be darker than the original. In order to correct this, the entire contrast is corrected in steps S56 and S58.

  In the second embodiment, brightness correction is performed in steps S56 and S58 in order to ensure brightness. However, an enlargement of the histogram that matches the peak positions of the histogram may be used here.

  In the present invention, the description has been made on the assumption that each pixel has RGB values, but the present invention can also be applied to a Bayer array used in a CCD.

Diagram of overall configuration of image processing apparatus Block diagram with a focus on control devices Main flowchart of image correction in the first embodiment Flow chart for executing image correction Flowchart of execution of image correction in the second embodiment Color cast detection flowchart Illustration of an example of the original histogram

Explanation of symbols

1 control device, 2 display, 201 CPU.

Claims (4)

Detection means for detecting the presence or absence of color cast on the color image data to be processed;
A certifying means for certifying a fog color for each pixel constituting the color image data detected as having color fog by the detecting means;
Of the color components consisting RGB of the pixel, and a subtraction means for performing subtraction processing to reduce the value of the color component corresponding to the head color, which is by Ri認 constant in the certification means,
The recognition means recognizes the color corresponding to the color component having the highest value among the color components composed of RGB of the pixel as a fog color,
The subtracting means sets the value of the color component corresponding to the fog color certified by the certification means among the color components composed of RGB of the pixel to be the same value as the value of the color component having the next largest value. Subtract,
Image processing device. The image processing apparatus according to claim 1 , further comprising a contrast correction unit that corrects a contrast of the color image data after the subtraction process by the subtraction unit. Detecting the presence or absence of color cast on the color image data to be processed;
Certifying a fog color for each pixel constituting color image data detected as having color fog ;
Wherein among the color components consisting of RGB pixel, the certified head computer readable recording medium storing a program for executing the steps on a computer to perform a subtraction process to reduce the value of the color component corresponding to the color Because
In the step of certifying the cover color, the program certifies a color corresponding to a color component having the highest value among color components composed of RGB of the pixel as a cover color,
In the step of performing the subtracting process, among the color components composed of RGB of the pixel, the value of the color component corresponding to the certified fog color is set to the same value as the value of the color component having the next largest value. Subtract,
As described above, the computer-readable recording medium causing the computer to execute the step of identifying the fog color and the step of performing the subtraction process. The computer-readable recording medium according to claim 3 , wherein the program further causes the computer to execute a step of correcting contrast of the color image data after the subtraction process in the step of performing the subtraction process.

Images