| 摘要 |
PURPOSE:To attain a simple and precise correction by executing a shading correction process, a matrix correction in which the color correction between a reference and plural sensors, etc., is executed and a masking process in which color correction for outputting a color printer, etc., is executed with a non-linear conversion and addition and subtraction. CONSTITUTION:As for the shading correction, a black reference signal and a white reference signal are added and averaged so as to be recorded in a line memory 110 before reading an original and the black reference signal is subtracted from a signal obtained by reading the original in an adder 113 and converted into a constant fold logarithmic conversion which is the non-linear conversion by a RAM 118 for non-linear conversion. Moreover, the non-linear conversion is executed to the white reference signal in the same way and a division process is executed with subtraction so as to execute the shading process. One-dimensional ROM or RAM table is used for the non-linear conversion. Since the process is executed only with the non-linear conversion and the subtraction, the process is simplified as compared with the process using an multiplier. Moreover, a matrix circuit or a masking circuit can be constituted by switching the constant value of the constant fold logarithmic conversion which is the non-linear conversion at every color separation signal.
|
