Color image processor

Details Color image processor Color image processor

1999-07-13

2003-05-06

Rogers, Scott (Department: 2624)

Facsimile and static presentation processing

Static presentation processing

Attribute control

C358S003230, C358S002100, C358S534000, C358S003270

Reexamination Certificate

active

06559976

ABSTRACT:

This application is based on application No. 10-197084/1998 filed in Japan, the content of which is incorporated hereinto by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processor which converts data of input color image to data used for image forming.
2. Description of the Prior Art
A color image processor receives image data of three primary colors of red, green and blue and converts them to cyan, magenta, yellow and black data used for image forming. The cyan, magenta, yellow and black data are subjected further to other processings and finally outputted to a printer to reproduce an image.
During the image data processing, conversion such as error diffusion is used for converting the original image data of a document to image data having a smaller number of pseudo-gradation levels. In the error diffusion, an error of input image data of read density at a pixel is diffused to adjacent pixels to conserve the image density of the document. The error diffusion has an advantage that a Moire pattern does not appear due to interaction of the frequencies of the document and the dither matrix, if compared with the dither method. Therefore, it is a powerful method of pseudo-gradation processing.
However, in the conventional bi-level error diffusion, image texture inherent in the error diffusion may happen, or nonsmoothness at character edges may be observed. Therefore, it is not appropriate for a color copying machine for which high quality is required. Then, multi-level error diffusion has been suggested for gradation representation at multi-levels equal to or higher than three levels.
However, multi-level error diffusion has following problems.
For error diffusion of n-level (n>2), (n−1) threshold levels are needed for quantization and in order to detect errors on the quantization, quantization error in each quantization level has to be detected. Therefore, a circuit therefor will have a large scale.
In the error diffusion, after the error is detected, an integrated value with adjacent pixels is determined, and the error is added to the input gradation value for the next pixel. Therefore, it is needed to process in the feedback system, including the error detection, the error integration and the addition, faster than the transmission route of the input pixel data. By adopting the n-level error diffusion (n>2), the processing time in the feedback system becomes long, and it is liable to exceed a limit to be operated by the circuit.
The threshold values of quantization are generally set uniformly in the gradation dynamic range of the input image data. However, as to color image, an image texture is liable to become noticeable due to color superposition of cyan, magenta, yellow and black. Further, a gradation range wherein the image texture is noticeable is affected by the gradation characteristics which depend on change in processes in the print engine for outputting an image.
The maximum level of quantization error in the multi-level error diffusion is smaller than that of the bi-level error diffusion, and the error convergence at character edges in an image is relatively fast. Then, non-smoothness at character edges is small. However, it is not satisfactory for a quality of a character image.
SUMMARY OF THE INVENTION
An object of the present invention is to provide an image processor which performs multi-level error diffusion of high quality with a simple structure.
In one aspect of the invention, an image processor according to the invention converts an m-bit image signal representing image density level of each pixel to an n-bit image signal where m and n are natural numbers. In the image processor, a first converter quantizes the m-bit image signal of pixels located at a character boundary to the n-bit image signal, while a second converter quantizes the m-bit image signal of pixels other than the pixels located at a character boundary to the n-bit image signal. Then, a selector synthesizes outputs of the first and second converters. In the second converter, a quantizer quantizes a first m-bit image signal of an object pixel to an n-bit image signal. An error detector detects an error generated in the quantization of the first m-bit image signal, and an adder adds the detected error to a next m-bit image signal. The error detector comprises a table which receives the output signal of the adder as an address and outputs data stored at the address as the error to the adder.
An advantage of the present invention is that error detection and error integration can be processed at a higher speed in multi-level error diffusion.


REFERENCES:
patent: 5130819 (1992-07-01), Ohta
patent: 5570432 (1996-10-01), Kojima
patent: 5708514 (1998-01-01), Higuchi et al.

Affiliated with

Also associated with

Rating

Color image processor does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Color image processor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Color image processor will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3093557