Facsimile and static presentation processing – Facsimile – Picture signal generator
Reexamination Certificate
2000-03-01
2003-05-20
Grant, II, Jerome (Department: 2624)
Facsimile and static presentation processing
Facsimile
Picture signal generator
C358S525000
Reexamination Certificate
active
06567192
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to enhancement of resolution of an image reading apparatus for reading an original image.
BACKGROUND OF THE INVENTION
An image reading apparatus is used as an image input unit for facsimile machines copier, computer, etc. The image reading method using a line image sensor having photoelectric converters arranged in a line is widely employed for its excellent total performance capability including the reading precision and speed. In a basic method, the line image sensor reads each line of the original image, and further moves to orthogonal direction (sub-scanning direction) to the line direction (main scanning direction) at intervals corresponding to the reading pixel interval in the line direction, thereby reading the image. On the other hand, in the background of sophistication of information processing and multimedia trend, a demand for enhancement of resolution of image reading apparatus is increased. Methods for enhancing the resolution of image reading apparatus include a method of increasing the precision of optical system such as the lens system and also increasing the density of photoelectric converters, and a method of reading different regions on the same line with the same line image sensor by changing over the optical system, but they are expensive because of the precise mounting procedure and complexity of mechanism.
Accordingly, it is proposed to increase the resolution by reading the image in the sub-scanning direction at the half line interval of the original. In this method, the reading density in the sub-scanning direction is double, but since the size of the pixels to be read is not changed, the resolution in the sub-scanning direction is increased but is not actually doubled. Further, in the main scanning direction, pixels are inserted between the pixels that have been already read, to make the pixel density equal in both main scanning and sub-scanning directions, and the value of the inserted pixels are given by the interpolating value of either adjacent pixel of the scanning direction or the average of both adjacent pixels. By this interpolation, the smoothness in the main scanning direction is improved, but the resolution is not increased.
In other method proposed to improve these points, every time the line image sensor is moved in the sub-scanning direction at the half line interval of the original, it is also moved in the main scanning direction reciprocally at the half pixel interval of the original, and the image pixels is read to obtain a checkered pixel plane, and the pixels in the interval are interpolated by the average of adjacent four pixels, so that the resolution is equalized in the main and sub-scanning directions. In this method, the resolution increased in the sub-scanning direction obtained by the above reading method is decreased in the sub-scanning direction and increased in the main scanning direction. In other words, the resolution once increased to double as the value in the ideal state in the sub-scanning direction is distributed only about {square root over ( )}2 times at most in the main scanning and sub-scanning directions.
The interpolation by the average used in both methods increases the smoothness of the image by nature, but blunts the contract in the edge of the image.
SUMMARY OF THE INVENTION
The invention presents a novel method to improve the resolution of the read image, and particularly reproduce edges of the original image at an excellent contrast, which is moving the line image sensor in the main scanning direction reciprocally at the half pixel interval of the original also every time scanning in the sub-scanning direction at the half line interval of the original, reading the image pixels, obtaining a checkered pixel plane of pixels being read, interpolating pixels of spaces in the said checkered pixel plane by utilizing the information of adjacent pixels to the pixels to be interpolated, and so in principle doubling the resolution in both main and sub-scanning directions.
Basically, plural interpolation candidate values for every pixel to be interpolated are calculated with the adjacent pixel values, and one of them is selected and interpolated.
One of its specific methods is a method of calculating the average values of all combinations of two pixels among the four adjacent pixels as the candidate values for every pixel to be interpolated, and selecting and interpolating the candidate value of the combination of two pixels having the minimum difference of the values.
In a modified method, alternatively, combinations of two pixels may be limited in two scanning directions.
For every pixel to be interpolated, in other method, when the direction of the two pixels having the minimum difference among all combinations of two elements of the four adjacent pixels, coincides with the two scanning directions, the candidate value determined from the values of the two pixels is selected, and when the direction of the two pixels differs from the scanning direction, the combination of two pixels is selected by referring to the value of the already interpolated pixel in the direction of the combination of the two pixels, and the candidate value determined from these values is selected and interpolated.
In a further different method, for every pixel to the interpolated, plural candidate values are determined from the values of the four adjacent pixels, and separately using values of several interpolated pixels adjacent to these four adjacent pixels, the predicted value for the pixel to be interpolated is determined, and the candidate value close to this predicted value is selected and interpolated.
When reading a color image, meanwhile, in all interpolating methods mentioned above, the candidate value of each pixel to be interpolated is selected according to the value of the luminosity to be combined from the values of the pixels of plural primary color images being read, or the values of the pixels of the green primary color images being read, and interpolated in every pixel of every color.
Moreover, to read each one of the pixels of the checkered pattern precisely, the voltage waveform for driving the actuator for moving the image sensor slightly in reciprocal motions must be kept as flat as possible at the reading position, and inclines steeply when moving so as to be connected smoothly, and its method is also disclosed.
For interpolation of pixels, a method of keeping the image sensor position always constant when starting image reading, and a method of obtaining a favorable checkered pixel plane also when resuming after interrupting reading are further disclosed.
The image reading apparatus of the invention employing these methods enhances the resolution of the read image, and reproduces the edges of the original image at excellent contrast in particular, and thereby contributes greatly to the progress of the information society, in particular, the multimedia society.
REFERENCES:
patent: 5418556 (1995-05-01), Andrews
patent: 2001/0046331 (2001-11-01), Altunbasak
patent: 60-16762 (1985-01-01), None
patent: 8-111774 (1996-04-01), None
patent: 9-214702 (1997-08-01), None
Kitagawa Shoichi
Mizusaki Masakazu
Narazaki Masanobu
Toyomura Yuji
Grant II Jerome
Worku Negussie
LandOfFree
Image reading apparatus 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 reading apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image reading apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3060582