Image analysis – Image enhancement or restoration
Reexamination Certificate
1998-10-13
2001-07-17
Lee, Thomas D. (Department: 2724)
Image analysis
Image enhancement or restoration
C358S406000, C358S474000, C358S494000
Reexamination Certificate
active
06263117
ABSTRACT:
BACKGROUND OF THE INVENTION
A. Field of the Invention
The present invention relates to an image calibration method for a contact type scanner, especially to an automatic image calibration method that solves the problems of missing line or skew image occurred when the sensor chips of the contact image sensor are not in alignment.
B. Description of the Prior Art
A conventional contact image sensor module (hereinafter referred to as CIS module) consists of multiple sensor chips that are perfectly aligned in a bar-shape module without any space therebetween. For instance, a 600 dpi CIS module usually consists of 15 sensor chips. And each sensor chip can read 340 pixels. These sensor chips in the CIS module shall be in perfect alignment as illustrated in FIG.
1
A. Accordingly, the image of a test chart read by these sensor chips will look like the image as illustrated in FIG.
2
A. However, when the sensor chips are not well-aligned during assemble, the phenomena of missing lines or skew images may occur as illustrated in FIGS.
1
B~
1
D. FIGS.
1
B~
1
C illustrate the situations when the sensor chips are not well-aligned in a vertical direction. On the other hand,
FIG. 1D
shows the situation when the sensor chips are not tightly connected together in a horizontal direction.
As illustrated in FIGS.
1
B~
1
C, when the sensor chips are not well-aligned in the vertical direction, the skew pixels can be detected from the horizontal lines or the slanting lines of the image of the test chart. On the other hand, when there is space between two neighboring sensor chips as illustrated in
FIG. 1D
, missing lines or pixels may be detectable from the edges of the slanting lines. In
FIG. 2D
, the missing pixels X, Y indicate the location where the sensor chips are not connected together tightly.
Although the quality of CIS modules from CIS manufactures is not always defected, a small or even invisible displacement of these sensor chips may affect the quality of images a lot. That explains why the image quality of the contact type scanners cannot be guaranteed. This has long been a problem for the scanner manufacturers because they have no control on the products what are from the third party.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide an automatic image calibration method that can calibrate various image problems caused by mal-aligned sensor chips based on three calibration parameters namely, a vertical displacement value, a horizontal displacement value, and a slope value.
It is another object of the present invention to provide an automatic image calibration method that provides the functions of fast image calibration and fine-tune image calibration, thereby a user can choose a function depending on the desired quality of images.
It is a further object of the present invention to provide an automatic image calibration method that is adaptable to various contact type scanners without having to digging out their possible hardware problems.
Accordingly, the computation of the calibration parameters can be implemented as an application program and executed in a personal computer. The application program is adaptable to various contact type scanners. At the step of system initiation, the inventive method involves in obtaining the calibration parameters for the contact type scanner being initialized. First, the application program reads the image of a test chart on the contact type scanner to compute the vertical displacement value, the horizontal displacement value, and the slope value for the contact type scanner. After the computation is finished, the calibration parameters will be stored in a memory which can be in the contact type scanner or in a host computer. After initialization, the application program can read and calibrate any image according to the calibration parameters stored in the memory.
The steps of computing the calibration parameters comprise: first, rotate the entire image to the same horizontal line based on the first pixel of the first sensor chip and the last pixel of the last sensor chip in the CIS module. Then, for each sensor chip, computing its vertical displacement value by subtracting the Y coordinate of the last pixel of the first sensor chip from the Y coordinate of the first pixel of the ith sensor chip. The vertical displacement value for each sensor chip indicates the sequence for an application program to read the image from a buffer. Thus, the vertical displacement value can eliminate skew images.
Then, based on the rotated image, for each sensor chip, computing its horizontal displacement value to its previous sensor chip by detecting and counting the number of missing pixels from the gray-levels variation of the slanting lines on the test chart. For each horizontal displacement value not equal to zero, fill in new pixels by the average value of the gray levels of the pixels neighboring to the missing pixels. The horizontal displacement value can eliminate the phenomena of missing lines occurred in an image.
The above mentioned calibration method can do coarse calibration at a fast speed. Nevertheless, for a high requirement of image quality, the slope value for each pixel must be computed. For each sensor chip, first compute its slope value by subtracting the Y coordinate of its first pixel from the Y coordinate of its last pixel. Then, after computing the slope value for each sensor chip, we can easily compute the slope value for each pixel by dividing each slope value by the number of pixels for each sensor chip. And then, we can calibrate each pixel of the sensor chip according to its slope value. This fine-tune calibration may take a longer time but with a best quality.
REFERENCES:
patent: 4870505 (1989-09-01), Mitsuki
patent: 5117295 (1992-05-01), Smitt
Lee Thomas D.
McDermott & Will & Emery
Umax Data Systems Inc.
LandOfFree
Automatic image calibration method for a contact type scanner does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automatic image calibration method for a contact type scanner, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automatic image calibration method for a contact type scanner will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2554274