Abstract: Sample areas are designated in a plurality of original images. Histograms representing density distribution for each color component are generated on the basis of the density signals for respective sample areas. Representative average density values in highlight-side and shadow-side in respective sample areas are determined from the histograms. A weighted average of respective representative average density values for the sample areas is calculated to obtain representative density values commonly applicable to the plurality of originals. The difference between the representative density values and the respective representative average density values are calculated, and, if the differences are within a predetermined range, reference density values for each color component are obtained on the basis of the common representative density values. The reference density values are used to define color separation rules common to all of the originals.

Abstract: An apparatus for converting an image signal representing an image having gradation includes a first density histogram generator. The first density histogram generator generates a first density histogram based on a first image signal which is obtained by reading a data read region. Of peak portions formed in the first density histogram, the density rank value of a peak portion which satisfies a predetermined condition is detected. A pixel-to-be-removed detector determines pixels-to-be-removed whose average densities are included in a pixel removal density range which corresponds to the detected peak portion density rank value. A second density histogram is generated from the pixels remaining after the removal of the pixels-to-be-removed. Based on the second density histogram, a cumulative density histogram is generated to thereby find a highlight point and a shadow point through which a gradation curve passes.

Abstract: A normalizing curve (C.sub.rx) is generated by correcting a reference normalizing curve. The character of the normalized curve depends on dummy shadow and highlight points (A.sub.4, A.sub.3) as well as shadow and highlight points (A.sub.2 A.sub.1), which are determined as a function of the statistical distribution of densities on an original image. Color original image signals representative of the original image are converted through the normalizing curve to obtain converted color image signals in which a gray balance in the intermediate density region is corrected.

Abstract: On a conversion plane of image data, first through third points (Q.sub.H, Q.sub.S, Q.sub.D) are provided. The first and second points are determined in accordance with highlight and shadow points of an original image, respectively, while the third point is determined in accordance with a dummy shadow point of the original image. A conversion curve (C.sub.a, C.sub.b) is so determined as to pass through the first and second points and to extend along a straight line (L) connecting the first and third points in the highlight region of the coordinate plane.

Abstract: A highlight point (HL.sub.R, HL.sub.G, HL.sub.B) and/or a shadow point (SD.sub.R, SD.sub.G, SD.sub.B) on a gradation correction curve are established in consideration of decoloration and color fog of an original. Density values of respective primary colors are obtained for each pixel on an original. Average density is computed for each pixel and a cumulative histogram of average density distribution is generated. Temporary reference density values corresponding to prescribed cumulative frequencies are found on the cumulative histogram. Reference color density values (D.sub.RH, D.sub.GH, D.sub.BH) are computed by averaging the density of each color for the pixels having average density between the temporary reference density and terminal density on the cumulative histogram. The highlight and shadow points are specified with one of the reference color density values.

Abstract: A color scanner capable of displaying a plurality of images in parallel comprises a scanning head for scanning a plurality of images for inputting image signals thereof; a first memory for storing image signals of the plurality of images inputted by the scanning head; a color correction circuit for correcting image signals stored in the first memory to set desired conditions for reproduction; an output skipping circuit for reducing the number of image signals corrected by the color correction circuit to a predetermined number so that they can be displayed on a color monitor; a second memory for storing image signals of the plurality of images skipped by the output skipping circuit; and a color monitor for simultaneously displaying on one screen the plurality of images corrected, based on the image signals stored in the second memory.