摘要 |
PURPOSE:To make it possible to execute fixed processing in each image by deciding an image to be diagnosed from the feature of the image which is an output from a detection mechanism and transmitting an extraction command to a feature extracting mechanism based upon the decided result. CONSTITUTION:The feature of an image 1 to be diagnosed is extracted by the feature extracting mechanism G6 and detected by the detection mechanism G14. The output of the mechanism 14 is inputted to an optimality deciding mechanism G13, which collates the input with plural stored patterns, determines the extraction quantity of the feature based upon similarity with the previously stored patterns and generates a command signal to the mechanism G6. Consequently, experienced operator's know-how can be adopted and the images of internal organs or the like can be automatically extracted. |