Television – Camera – system and detail – Combined image signal generator and general image signal...
Reexamination Certificate
1998-01-26
2004-10-19
Garber, Wendy R. (Department: 2612)
Television
Camera, system and detail
Combined image signal generator and general image signal...
C348S222100, C382S289000, C358S519000
Reexamination Certificate
active
06806903
ABSTRACT:
This application is based on patent application Nos. 9-12999, 9-13000, 9-13001, 9-13002, 9-13003, 9-13004, 9-13005, 9-13006, 9-13019, 9-13020, and 9-13021 filed in Japan, the contents of which is hereby incorporated by reference.
BACKGROUND OF THE INVENTION
This invention relates to an image capturing apparatus capable of picking up a light image of an object by an image pickup device such as CCD (Charge Coupled Device) by photoelectrically converting it into an electric image and storing it in a storage medium such as a hard disk card after applying a specified image processing thereto.
There have been known electronic cameras as image capturing apparatus. Electronic cameras have, as compared with conventional cameras which record light images on silver halide film, an advantage of picking up images of satisfactory quality by suitably applying an image quality processing to the picked up images according to image capturing purposes and types of objects since the quality of captured image can freely be processed. Thus, the digital cameras are used as devices not only for usual picture taking but also for image capturing images such as characters and figures drawn on a white board at, e.g., a conference hall.
In the case that a white board on which characters, figures, etc. are drawn by an electronic camera, the main purpose of the image capturing is to record a representation represented by characters, figures and the like on the white board. Accordingly, it is desirable to apply such a &ggr;-correction to the captured image as to enhance the clearness of a representation portion (a portion of characters and figures) by making a white portion (white board portion) white. In this case, since a variation in the character density on the white board and an illuminance nonuniformity are large, it is desirable to correct the illuminance nonuniformity (shading correction) by dividing the captured image into a plurality of blocks in a two-dimensional manner and applying the &ggr;-correction block by block.
Specifically, if the white board is assumed to be illuminated by ceiling lights of the room and sunlight coming through the windows, illuminance nonuniformity occurs due to a nonuniform illumination light. By the multiplying effect of this illuminance nonuniformity and a distribution of incident light amount by the so-called law of Cos
4
&thgr; according to which an image at an off-axis object point which is incident on the entrance pupil of the taking lens at an angle &ohgr;, a distribution of the output of the image pickup devices such as CCDs largely varies along horizontal and vertical directions on a sensing surface.
Thus, it is desirable to perform the illuminance nonuniformity correction by dividing the picked image into a plurality of blocks in a two-dimensional manner and by applying the &ggr;-correction according to the illuminance in the block for each block. More preferably, it is preferable to set a proper &ggr;-characteristic for each block by making the block size as small as possible in order to avoid the creation of a pseudo line at the boundary of the blocks resulting from a sudden change in the &ggr;-characteristic in the case that the &ggr;-characteristic set for each block largely changes between neighboring blocks.
The &ggr;-characteristic for each block used for the &ggr;-correction performed block by block can be set using the histogram of level-frequency distribution of the pixel data included, for example, in the block. In other words, when an image of characters, figures or the like drawn on the white board is picked up and a histogram of level-frequency distribution of pixel data constituting the picked image is generated, the generated histogram of level-frequency distribution is normally a two-peak distribution histogram having a convex portion corresponding to the character portion at a dark side and a convex portion corresponding to the white portion (white board) at a bright side. The white level is detected from the convex portion corresponding to the white portion, and the &ggr;-characteristic is so set as to convert the pixel data above this white level into pixel data of a predetermined white saturation level.
According to the &ggr;-characteristic setting method using the histogram of level-frequency distribution, the set &ggr;-characteristic varies according to the number and size of characters included in the block in the case that the picked image is divided into rectangular blocks. If the block size in relation to the character size is improper, a suitable &ggr;-characteristic cannot be obtained. For example, if the block size is considerably smaller than a suitable size in relation to the character size, the character portion takes up a large area in the block and an area of the white portion is small. The convex portion corresponding to the white portion of the histogram of level-frequency distribution generated using the pixel data in the block becomes small and it is difficult to determine the white saturation level based on this convex portion. Conversely, if the block size is considerably larger than the suitable size, the convex portion corresponding to the white portion of the histogram of level-frequency distribution generated using the pixel data in the block is sufficiently large. However, since the white portion takes up a large area in the block, the convex portion corresponding to the white portion becomes moderately sloped due to the influence of the nonuniform illuminance. Thus, it is difficult to stably set the white saturation level based on this convex portion.
The character size in a field of the viewfinder easily varies according to the object distance and the image capturing magnification in picture taking. However, it is not preferable that the quality of images obtained by image capturing the same object considerably changes according to the object distance and the image capturing magnification. Accordingly, the block size needs to be set at the specified size in relation to the character size so that a suitable histogram of level-frequency distribution can be obtained.
Further according to the &ggr;-characteristic setting method using the histogram of level-frequency distribution, the histogram of level-frequency distribution of the block including both the white board portion and the background portion displays a two-peak distribution having a convex portion corresponding to the white board portion and a convex portion corresponding to the background portion in a white area. Thus, there is a likelihood that the white level is erroneously detected based on the convex portion corresponding to the background portion.
In the case that the background portion is brighter than the white board portion, the white level is detected based on the convex portion corresponding to the background portion and the &ggr;-characteristic is set using this white level, the pixel data of the white board portion are not converted into pixel data of specified saturation level (pure white) in monochromatic image capturing. If the &ggr;-correction is performed to intensify the black portion to emphasize the characters, the pixel data of the white board portion are converted into black in some cases. This leads to a disadvantage that the white board portion of the block including the background portion turns black. In the case of a color image, if the &ggr;-characteristic for the image of any color components is set as above, a part of color components are completely converted into those of the black saturation level. Therefore, a chromatic coloring phenomenon occurs in the white board portion.
In the case that the white board is captured together with its background, the image quality is considerably reduced, making the image hard to be seen if the coloring phenomenon occurs in the white board portion in a boundary area between the white board portion and the background portion during the image processing, namely the illuminance nonuniformity correction. Thus, it is desirable to detect the boundary area between the white board portion and the
Fuji Shinichi
Okisu Noriyuki
Minolta Co. , Ltd.
Rosendale Matthew L
LandOfFree
Image capturing apparatus having a &ggr;-characteristic... 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 capturing apparatus having a &ggr;-characteristic..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image capturing apparatus having a &ggr;-characteristic... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3308719