Image processing system and image processing method

An image processing method comprising:

• an input step of inputting an image;
• a first extraction step on a processor of performing a first region extraction process to the image input in the input step to extract a first region group including one or more regions;
• and a second extraction step on the processor of performing a second region extraction process for every region included in the first region group,
• wherein in the first extraction step, the region is extracted on the basis of a logical operation of a first binary image and a second binary image, the first binary image being based on a result of first filtering with first pass frequency band characteristic, the second binary image being based on a result of second filtering with second pass frequency band characteristic.

An image processing method comprising:

• a filtering step of performing first and second band pass filtering to an image;
• a first binary image formation step on a processor of forming first and second binary images from the first and second band pass filtering results with first and second frequency band characteristics;
• region group specification step on the processor of specifying a first region group including a structural component to be extracted in the image and a second region group including no structural component on the basis of logical operation for the first and second binary images;
• a re-extraction step on the processor of again extracting a desired structural component from the first region group on the basis of the second region group specified in the region group specification step;
• and
• a second binary image formation step on the processor of forming a binary image every region included in the first region group.

An image processing method comprising:

• an input step of inputting an image;
• an extraction step of extracting a predetermined structural component from the image input in the input step;
• and
• a feature-amount calculation step on a processor of calculating an amount of feature based on a width of the structural component extracted in the extraction step,
• wherein the extraction in the extraction step includes a binary image formation step of forming a binary image of the predetermined structural component in the image,
• the width of the structural component is based on the result of a distance transform and skeleton process for the binary image, and
• the amount of feature includes a numeric value to evaluate the magnitude of variation in the width of the structural component.

An image processing method comprising:

• an input step of inputting an image comprising a plurality of color signals;
• an extraction step of extracting desired structural components from at least two of the color signals constituting the image input in the input step;
• and
• a calculation step on a processor of calculating the combination of amounts of feature based on the desired structural components extracted in the extraction step,
• wherein the extraction in the extraction step includes a binary image formation step of forming a binary image of the predetermined structural component in the image,
• the width of the structural component is based on the result of a distance transform and skeleton process for the binary image, and
• the amount of feature includes a numeric value to evaluate the magnitude of variation in the width of the structural component.

An image processing method comprising:

• an input step of inputting an image including a plurality of color signals;
• an extraction step of extracting desired structural components from at least two of the color signals constituting the image input in the input step;
• a combination step on a processor of combining the structural components extracted in the extraction step;
• and
• a calculation step of calculating the amount of feature based on the combination result in the combination step,
• wherein the extraction in the extraction step includes a binary image formation step of forming a binary image of the predetermined structural component in the image,
• the width of the structural component is based on the result of a distance transform and skeleton process for the binary image, and
• the amount of feature includes a numeric value to evaluate the magnitude of variation in the width of the structural component.

An image processing system comprising:

• an input unit for inputting an image;
• an extraction unit for performing matching between a template, obtained by modeling a predetermined structural component in the image, and the image input in the input unit to extract a structural component serving as an extraction target in the image;
• and
• an evaluation unit for setting a plurality of different templates of the structural components to evaluate a result of matching using the different templates, wherein
• the structural component is extracted on the basis of the evaluation result in the evaluation unit, and each of the different templates is determined on the basis of the number of pixels corresponding to a width of the structural component and a magnitude of variation in the width.