JP2007078864A - Method for eliminating color unevenness of optical microscope photograph - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for eliminating color unevenness of an optical microscope photograph enabling high image analysis by removing the various color unevenness caused on the optical microscope photograph. <P>SOLUTION: The method is employed to eliminate the color unevenness by finding an average density value of colors of the whole image concerning each component image of R, G and B of a color image of the optical microscope photograph, finding an average of color density concerning a pixel within a prescribed level to the average density value of the colors of the whole image by a density level at each line of an X direction and a Y direction, finding a difference between the average density value of the colors of an attention line and the average density value of the colors of the whole image, adding the difference in color density of each pixel of the noticed line to perform flattening processing of the X direction and the Y direction, and replacing each component image of flattening processed R, G and B with each component image of original R, G and B. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The invention of this application relates to a method for removing uneven color in a photograph taken with an optical microscope such as a color laser microscope or a metal microscope.

  When using an optical microscope such as a color laser microscope or a metal microscope to observe the form of corrosion or cracking of metal materials, etc., color unevenness occurs in the optical microscope photograph due to subtle defects in the polished state, and the accuracy of the photograph decreases. There are many things.

  There is shading correction as a method for correcting color unevenness in black and white or color images. This shading correction is a method for correcting an image by setting a color profile related to the apparatus (for example, Patent Documents 1 and 2).

  However, in shading correction, it is possible to correct the color unevenness inherent in the optical microscope, but it is impossible to remove the color unevenness caused by the polished state as in the polished sample as described above.

As described above, the conventional technique cannot cope with the removal of various color unevenness generated in the optical micrograph.
JP 2005-94214 A JP 2005-109831 A

  Therefore, the invention of this application was made in view of the circumstances as described above, and removes various color unevenness occurring in the optical micrograph, and removes the color unevenness of the optical micrograph that enables advanced image analysis. It is an object to provide a method.

  The invention of this application solves the above-mentioned problem. First, for each of the R, G, and B component images of the color image of the optical micrograph, the average density value of the color of the entire image is obtained, and the X direction For each line, the average of the color density is obtained for pixels whose density level is within a predetermined level with respect to the average density value of the color of the entire image, and the average density value of the color of the target line and the average density value of the color of the entire image are obtained. Is obtained, and the difference is added to the color density of each pixel of the line of interest to perform the flattening process in the X direction. Then, for each line in the Y direction, the density level is the average density of the colors of the entire image. The average of the color density is obtained for pixels within a predetermined level with respect to the value, the difference between the average density value of the color of the target line and the average density value of the color of the entire image is obtained, and the difference is calculated as the color density of each pixel of the target line. By adding Y direction An optical microscope photograph characterized by performing color flattening and removing color unevenness by replacing the flattened R, G, B component images with the original R, G, B component images. A method for removing color unevenness is provided.

  According to a second aspect of the present invention, there is provided a method for removing color unevenness in an optical micrograph, wherein the color image of the optical micrograph is a 24-bit color image of 8 bits each for R, G, and B. provide.

Thirdly, in the first or second invention described above, color unevenness in an optical micrograph, wherein a mask is applied to a specific area of the color image to perform flattening processing in the X direction and the Y direction. A removal method is provided.

  According to the invention of this application, it is possible to remove various color unevenness generated in an optical micrograph as in the case of imaging a polished metal surface, and an image when observing corrosion or cracking of a metal material or the like. It is possible to improve the accuracy of analysis.

  The invention of this application has the features as described above, and an embodiment thereof will be described below.

  In the method for removing color unevenness of an optical micrograph according to the invention of this application, for each of the R, G, and B component images of the color image of the optical micrograph, an average density value of the color of the entire image is obtained and flattened in the X direction. After the processing, the Y-direction flattening process is performed, and the unevenness of color is removed by replacing the flattened R, G, and B component images with the original R, G, and B component images. .

  In the flattening process in the X direction, for each line in the X direction, an average color density is obtained for pixels whose density level is within a predetermined level with respect to the average density value of the color of the entire image, and the average density of the color of the line of interest The difference between the value and the average density value of the color of the entire image is obtained, and the difference is added to the color density of each pixel of the line of interest.

  In the Y-direction flattening process, for each line in the Y-direction, an average color density is obtained for pixels whose density level is within a predetermined level with respect to the average density value of the colors of the entire image, and the average density of the color of the line of interest The difference between the value and the average density value of the color of the entire image is obtained, and the difference is added to the color density of each pixel of the line of interest. Hereinafter, it demonstrates in detail, showing a specific example.

  FIG. 1 is a flowchart showing a procedure of a method for removing color unevenness in an optical micrograph according to an embodiment of the present invention. 2 is an image of a typical color optical micrograph with uneven color (stainless steel corroded by magnesium chloride droplets), FIG. 3 is an R component image of FIG. 2, and FIG. 4 is uneven color. Removed R component image, FIG. 5 shows an image from which color unevenness has been removed for all R, G, and B components.

  Here, a case where color unevenness is removed from a 24-bit color image of 8 bits each for R, G, and B will be described as an example, but the invention of this application is not limited to this. Color unevenness removal processing can be performed using commercially available image processing software that can decompose a 24-bit RGB image into 8-bit R, G, and B component images and synthesize them into a 24-bit RGB image.

  First, 24-bit color digital image information of an optical microscope photograph taken by a CCD camera or the like attached to the optical microscope is input to a processing device such as a personal computer equipped with the image processing software (step S1).

  Next, a mask (portion surrounded by a circle in FIG. 3) is applied to the polished surface (a specific surface) of the color image, and a portion not covered with the mask is set as a processing target (step S2). In this case, the mask is applied because only information on a gradual color density change of the polished surface is acquired. If all color information is captured without masking, flattening cannot be performed.

Next, for each of the R, G, and B component images of the color image to be processed, the average density value of the color of the entire image is obtained (step S3). For each 8-bit component image of R, G, and B, the density level is within a predetermined level with respect to the average density value for each line in the X direction (for example, the density level is within 30 with respect to the average density value of the entire pixel). ) For the pixels of () is obtained (step S4). Next, a difference between the average density value of the color of the target line and the average density value of the color of the entire image is obtained, and the difference is added to the color density of each pixel of the target line (step S5). As described above, the flattening process in the X direction is performed, and an image having no color unevenness in the Y direction is obtained.

  Next, for each of the R, G, and B component images, for each line in the Y direction, an average color density is obtained for pixels whose density level is within a predetermined level (same as above) with respect to the average density value (step S6). ). Then, the difference between the average density value of the color of the target line and the average density value of the color of the entire image is obtained, and the difference is added to the color density of each pixel of the target line (step S7). As described above, the flattening process in the Y direction is performed, and an image having no color unevenness in the X direction is obtained.

  When the R, G, and B component images have been flattened in the X and Y directions, the flattened R, G, and B component images are replaced with the original R, G, and B component images. Then, a color image having no color unevenness as a whole as shown in FIG. 5 is obtained (step S8).

  Although omitted in the above description, the average color density of the R, G, and B component images after the flattening process is the same as the average color density of the original R, G, and B component images obtained in advance. It is preferable to correct so.

It is a flowchart which shows the procedure of the color nonuniformity removal method of the optical micrograph which concerns on one Embodiment of invention of this application. It is a figure which shows the image (stainless steel corroded by adhering a magnesium chloride droplet) of the color optical micrograph with typical color irregularity. It is a figure which shows the R component image of FIG. It is a figure which shows R component image from which the color nonuniformity was removed. It is a figure which shows the image which removed the color nonuniformity about R, G, and B component.

Claims (3)

  For each of the R, G, and B component images of the color image of the optical micrograph, the average density value of the color of the entire image is obtained, and for each line in the X direction, the density level is relative to the average density value of the color of the entire image. Calculating the average color density of pixels within a predetermined level, obtaining the difference between the average density value of the color of the target line and the average density value of the color of the entire image, and adding the difference to the color density of each pixel of the target line. Then, the flattening process in the X direction is performed, and then, for each line in the Y direction, the average of the color density is obtained for pixels whose density level is within a predetermined level with respect to the average density value of the color of the entire image, and the line of interest The difference between the average density value of the color of the image and the average density value of the color of the entire image is obtained, and by adding the difference to the color density of each pixel of the line of interest, the Y-direction flattening process is performed and the flattening process is performed. R, G, B component images Original R, G, uneven color removal method of the optical micrograph, and removing the color unevenness by replacing each component image of B.   2. The method according to claim 1, wherein the color image of the optical micrograph is a 24-bit color image of 8 bits each for R, G, and B.   3. The method for removing color unevenness in an optical micrograph according to claim 1, wherein a flattening process in the X direction and the Y direction is performed by applying a mask to a specific area of the color image.