Image analysis – Color image processing
Reexamination Certificate
2000-03-10
2004-03-02
Johns, Andrew W. (Department: 2621)
Image analysis
Color image processing
C382S165000, C382S167000, C358S518000, C358S520000
Reexamination Certificate
active
06701007
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing device, an image processing method, and a recording medium, and in particular, to an image processing method in which the gray balance of an original image recorded on a photographic film is estimated, an image processing device to which this image processing method is applied, and to a recording medium on which is recorded a program for operating a computer as the image processing device.
2. Description of the Related Art
The color balance of an image which is made visible on a photographic film by carrying out processings such as development for a photographic film with an object photographed and recorded thereon by a camera or the like (hereinafter simply referred to as an image recorded on the photographic film) is influenced by characteristics of the photographic film itself, processing conditions for developing processing, and the like (hereinafter generically referred to as film characteristics) and is deviated or biased from the color balance of the object at the time of photographing. Therefore, when the image recorded on the photographic film is recorded onto a recording material such as a photographic printing paper or is displayed on display means such as a display, the color balance of the image to be recorded onto the recording material or the image to be displayed on the display means (hereinafter generically referred to as the output images) must be corrected such that the color balance of the output image corresponds to the color balance of the object at the time of photographing (such that gray portions of the object at the time of photographing are reproduced as gray in the output image).
As an example of methods of correcting the color balance of output images, a method is known in which pixels corresponding to highlight regions of an image (i.e., an original image) recorded on a photographic film (e.g., pixels having a maximum density in a negative image, and pixels having a minimum density in a positive image) are assumed to be white, while pixels corresponding to shadow regions of the original image (e.g., pixels having a minimum density in the negative image, and pixels having a maximum density in the positive image) are assumed to be black. The gray balance representing the color balance, in the original image, of gray portions of the object at the time of photographing is estimated, and the color balance of an output image is corrected on the basis of the estimated gray balance. In this method, pixels corresponding to highlight regions and pixels corresponding to shadow regions are respectively extracted from the original image. Then, for example, an axis which connects, by a straight line, the points on an RGB density coordinate system which correspond to both of the pixels is determined as the gray axis representing the gray balance.
However, in the aforementioned correction method, when the color of the pixels corresponding to the highlight regions in the original image is not white, as can be seen, for example, in an image in which a person has been photographed by using an electronic flash and cheek portions of the person's face (i.e., flesh-colored portions) are highlighted, the gray axis representing the appropriate gray balance cannot be determined (the same holds for the shadow regions). Further, the color balance of the output image deviates to a complementary color for the color of the pixels corresponding to the highlight regions. This deviation is referred to as “highlight failure”. The frequency of generation of an original image susceptible to the above-mentioned highlight failure is relatively high, and a drawback arises in that the probability of obtaining an output image with an appropriate color balance is low.
Further, a method is also known in which, on the basis of Evans' theory, an image having a constant LATD (i.e., light accumulated transmission density) for each of component colors (e.g., R, G, and B) is regarded as an image whose color balance corresponds to the color balance of an object at the time of photographing (i.e., an image having an appropriate gray balance), and the LATD for each of the component colors (e.g., R, G, and B) of the original image is measured. The color balance of an output image is corrected so that the LATD for each of the component colors of the output image becomes constant.
However, in the aforementioned correction method, in a case in which a non-gray region having a substantially constant hue (e.g., a region corresponding to green turf, blue sky, blue sea, or the like) occupies a relatively wide area of the original image, when correction is carried out so that the LATD for each of the component colors becomes constant, the color balance of the output image deviates to the complementary color of the aforementioned region (this deviation is referred to as “color failure”). The frequency of generation of an original image susceptible to the above-described color failure is relatively high. Therefore, a drawback arises in that, as in the case with the other correction method described above, the probability of obtaining an output image with an appropriate color balance is low.
Moreover, Japanese Patent Application Laid-Open (JP-A) No. 9-83825 discloses a linear transformation technique. In this technique, image data DR′, DG′, and DB′ regarding low saturation pixels is acquired from image data DR, DG, and DB representing an original image. Shadow points DRs, DGs, and DBs and highlight points DRh, DGh, and DBh in the image data DR′, DG′, and DB′ are respectively determined. Regarding one of a pair of the image data (DR′, DG′) corresponding to each other for each pixel, values in the other are averaged for each of the same values in the one to obtain a set of image data (DR″, DG″). Subsequently, a relationship between the densities of the two colors (or R, G) is determined from the set of the image data (DR″, DG″), the shadow points (DRs, DGs), and the highlight points (DRh, DGh). On the basis of this relationship, linear transformation is carried out so that at least one of the image data DR and DG is made equal to the other over the entire region thereof.
By excluding data of high saturation pixels from the image data representing the original image from the object of computation, the above-described technique is meant to reduce effects caused by the original image susceptible to color failure. In addition to the high saturation pixels, pixels which are adjacent to the high saturation pixels and whose hue difference with respect to the high saturation pixels is within a predetermined range of values, are also excluded from the object of computation. However, in practice, even if the above-mentioned processes are effected with respect to the original image susceptible to color failure, among the pixels of non-gray regions having a substantially constant hue which occupy a relatively wide area of the original image, many pixels remain without being excluded from the object of computation. Therefore, a drawback arises in that correction accuracy of the color balance for the output image is not sufficient when the original image is susceptible to color failure.
Further, JP-A No. 5-297473 discloses a technique in which a large volume of image information obtained by reading images recorded on various types of photographic films is accumulated and stored for each type of the photographic film, and when an image recorded on a particular type of photographic film is to be recorded by exposing a photographic printing paper, exposure conditions are determined by regarding, as a representation of the gray balance, the color balance corresponding to the average of the image information accumulated and stored with regard to the particular film type.
In the above technique, however, the accuracy of the color balance corresponding to the average of the accumulated and stored image information (i.e., deviation with respect to
Alavi Amir
Fuji Photo Film Co. , Ltd.
Johns Andrew W.
LandOfFree
Image processing device, image processing method, and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing device, image processing method, and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing device, image processing method, and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3225584