EP0198481A2 — Image processing apparatus

Claim 1

Independent Claim 100 words

An image processing method for extracting (12) an area whose density distribution characteristic meets a predetermined condition from an input image (11) of multiple gray levels and extracting an object area as an output image (13) in accordance with information on an extractn area, comprising the steps of: (a) designating (15) a limited area having a substantially uniform density distribution and containing the object area, to said input image; and (b) extracting a density distribution characteristic of the input image in the designated area and determining (16) a condition for the area extraction in accordance with the density distribution characteristic.

Claim 2

Dependent Claim 100 words

An image processing method for extracting (12) an area whose density distribution characteristic meets a predetermined condition from an input image (11) of multiple gray levels and extracting an object area as an output image (13) in accordance with information on an extractn area, comprising the steps of: (a) designating (15) a limited area having a substantially uniform density distribution and containing the object area, to said input image; and (b) extracting a density distribution characteristic of the input image in the designated area and determining (16) a condition for the area extraction in accordance with the density distribution characteristic.

Claim 3

Dependent Claim 25 words

An image processing method according to Claim 1 wherein said designated area includes an object area candidate area or said area and a background area.

Claim 4

Dependent Claim 35 words

An image processing method according to Claim 1 further comprising a step of reading in the input images a plurality of times and extracting the density distribution characteristics of the input in said designated area.

Claim 5

Dependent Claim 86 words

An image processing method according to Claim 1 wherein said step of extracting the characteristic includes substeps of normalizing a density-frequency distribution of the input image such that the density of the input image coincides with a density at a characteristic point in a preset density-frequency distribution of a reference image, setting a characteristic amount of the density distribution in a partial image around pixels of the normalized image as a new density of said pixels, and binarizing said new density by a preset threshold level.

Claim 6

Dependent Claim 33 words

An image processing method according to Claim 1 wherein said predetermined condition to be met by the density distribution characteristic includes a binarization threshold and a size and a type of local filtering.

Claim 7

Dependent Claim 158 words

An interactive image processing method for storing image characteristics of object areas and procedures for extracting the object areas, selecting the procedure for extracting the object area by comparing the image characteristic of an externally applied input image with the stored image characteristics, and externally inquiring for inputting the image characteristic of the input image in selecting the object area extraction procedure, comprising the steps of: (a) designating a limited area having a substantially uniform characteristic and containing at least the object area or candidate area, to the input image; (b) determining the object area extraction condition information in accordance with the image characteristic extracted form the object area presented to said limited area; and (c) evaluating the input image characteristic in accordance with the extraction condition information and a predetermined procedure; (d) the object area extraction procedure being derived from the input image characteristic in accordance with the results of said inquiry and said extracted image characteristic.

Claim 8

Dependent Claim 158 words

An interactive image processing method for storing image characteristics of object areas and procedures for extracting the object areas, selecting the procedure for extracting the object area by comparing the image characteristic of an externally applied input image with the stored image characteristics, and externally inquiring for inputting the image characteristic of the input image in selecting the object area extraction procedure, comprising the steps of: (a) designating a limited area having a substantially uniform characteristic and containing at least the object area or candidate area, to the input image; (b) determining the object area extraction condition information in accordance with the image characteristic extracted form the object area presented to said limited area; and (c) evaluating the input image characteristic in accordance with the extraction condition information and a predetermined procedure; (d) the object area extraction procedure being derived from the input image characteristic in accordance with the results of said inquiry and said extracted image characteristic.

Claim 9

Dependent Claim 40 words

An interactive image processing method according to Claim 6 wherein said steps of designating the limited area designates the limited area having the substantially uniform density distribution and containing the object area from the input image of multiple gray levels.

Claim 10

Dependent Claim 39 words

An interactive image processing method according to Claim 7 wherein said extraction condition information includes a binarization threshold and a size and a type of local filtering as a condition to be met by the extracted density distribution characteristic.

Claim 11

Dependent Claim 25 words

An interactive image processing method according to Claim 7 wherein the image characteristic as said extraction condition information includes density distribution characteristic and geometrical characteristic.

Claim 12

Dependent Claim 102 words

An image processing apparatus comprising: segmentation means (12, 24) for extracting an area whose density distribution characteristic meets a predetermined condition, from an input image (11) of multiple gray levels; said segmentation means extracting an object area as an output image (13) from said input image; area designation means (15, 25) for designating a limited or localized area having substantially uniform density distribution and containing the object area, to said input image; and condition determination means (16, 26 -28) for determining the condition of said segmentation means in accordance with the density distribution characteristic of the input image in said designated area.

Claim 13

Dependent Claim 102 words

An image processing apparatus comprising: segmentation means (12, 24) for extracting an area whose density distribution characteristic meets a predetermined condition, from an input image (11) of multiple gray levels; said segmentation means extracting an object area as an output image (13) from said input image; area designation means (15, 25) for designating a limited or localized area having substantially uniform density distribution and containing the object area, to said input image; and condition determination means (16, 26 -28) for determining the condition of said segmentation means in accordance with the density distribution characteristic of the input image in said designated area.

Claim 14

Dependent Claim 18 words

An image processing apparatus according to Claim 10 wherein said area designation means includes a coordinate input device.

Claim 15

Dependent Claim 90 words

An image processing apparatus according to Claim 10 wherein said segmentation means includes a density-frequency distribution normalization unit for normalizing the distribution such that the density of the input image coincides with the density at a characteristic point in a predetermined density-frequency distribution of a reference image, a partial density characteristic operation unit (27) for storing a density distribution characteristic in the partial image around the normalized image pixels as a new density of said pixels, and a binarization unit (23) for binarizing the input image with a predetermined threshod.

Claim 16

Dependent Claim 132 words

An image processing apparatus according to Claim 12 wherein the density at the characteristic point in the density-frequency distribution used in said density-frequency distribution normalization unit is determined by calculating density-frequency distributions per unit area (72, 73) of the object area and the background in the designated area of the reference image, calculating density-frequency distributions per unit area (74, 75) of the object areas and the backgrounds in the designated areas of a plurality of normalized images of the plurality of input images, calculating sums (76) of deviations between above distributions for the object area and the background, respectively, and determining the density (78) at a point which results in a minimum accumulation of the sum of the deviations for the object area and the sum of the deviations for the background.

Claim 17

Dependent Claim 142 words

An image processing apparatus according to Claim 12 wherein the density distribution characteristic in the partial image used by said partial density characteristic operation unit is determined by performing the partial characteristic operations to a plurality of input images, binarizing the input images by a predetermined threshold to produce the plurality of binary images, calculating a sum of ratios of areas of the binary iamges contained in the object areas in the designated areas and the asreas of the designated areas, calculating a sum of ratios of areas of backgrounds in the designated areas minus the areas of the respective binary image areas contained in the backgrounds and the areas of the background, and determining a characteristic which results in a maximum accumulation of a sum of the ratios for the backgrounds and a sum of the ratios for the object areas.

Claim 18

Dependent Claim 150 words

An image processing apparatus according to Claim 12 wherein the threshold used in said binarization unit (98) is determined by preparing a plurality of threshold calculation means (111, 111', 111"), calculating density-frequency distributions per unit area of the object areas and the background in the designated areas for the images derived from a plurality of input images by performing the partial density characteristic operations, calculating total density-frequency distributions which are sums for the same components of the density-frequency distributions for the object areas and the backgrounds, and selecting a threshold which results in a maximum accumulation of the sum frequency of the total density-frequency distributions of the object areas in the density range segmented by the threshold and a sum frequency of the total density-frequency distribution of the backgrounds outside of said density range, from the thresholds for binarizing the object areas calculated by said plurality of threshold calculation means.

Claim 19

Dependent Claim 36 words

An image processing apparatus according to Claim 10 wherein said condition determination means (16, 26) determines the condition in accordance with the density distribution characteristics of the object area and the background in the designated area.

Claim 20

Dependent Claim 37 words

An image processing apparatus according to Claim 10 wherein said area designation means (15, 25) includes an image display screen and means for designating the area by painting only the object area candidate area on the screen.

Claim 21

Dependent Claim 22 words

An image processing apparatus acording to Claim 17 wherein said means for designating the area by painting includes an edge emphasizing module.

Claim 22

Dependent Claim 30 words

An image processing apparatus according to Claim 10 wherein said area designation means includes a detection module for extracting a density distribution of an image only in the designated area.

Claim 23

Dependent Claim 149 words

An interactive image processing system comprising; knowledge memory means (221) for storing a plurality of sets of object image characteristics and procedures for extracting the object areas; deduction means (226) for comparing an image characteristic of an extemally applied input image and the image characteristics stored in said knowledge memory means to extract the corresponding object area; inquiry means (228) for extemally inquiring to enter the image characteristic of the input iamge to said deduction means; area designation means for designating the object area; and image characteristic evaluation means (227) for evaluating the image characteristic in accordance with the predetermined procedures based on the designated area information derived from said area designation means; the results of said inquiry means and said image characteristic evaluation means being applied to said deduction means to determine the conditions to extract the object area based on the image characteristic of the input image.

Claim 24

Dependent Claim 149 words

An interactive image processing system comprising; knowledge memory means (221) for storing a plurality of sets of object image characteristics and procedures for extracting the object areas; deduction means (226) for comparing an image characteristic of an extemally applied input image and the image characteristics stored in said knowledge memory means to extract the corresponding object area; inquiry means (228) for extemally inquiring to enter the image characteristic of the input iamge to said deduction means; area designation means for designating the object area; and image characteristic evaluation means (227) for evaluating the image characteristic in accordance with the predetermined procedures based on the designated area information derived from said area designation means; the results of said inquiry means and said image characteristic evaluation means being applied to said deduction means to determine the conditions to extract the object area based on the image characteristic of the input image.

Claim 25

Dependent Claim 22 words

An interactive image processing system according to Claim 20 wherein said image characteristic includes a density distribution characteristic and a geometrical characteristic.

Claim 26

Dependent Claim 43 words

An interactive image processing system according to Claim 20 wherein said inquiry means inquires in a form of message in accordance with the image characteristic stored in said knowledge memory and supplies the image characteristic of the input image to said deduction means.

Claim 27

Dependent Claim 34 words

An interactive image processing system according to Claim 20 wherein said area designation means include a coordinate input device to designate an area coordinate of the coordiante input device corresponding to the input image.

Image processing apparatus

Abstract

Inventors

Bibliographic Details

Claims Analysis

References Cited (Backward Citations)

USPTO Annuity & Maintenance Windows

Assigned Categories & Tags

Assign Category Tag