Claims Analysis
Total Claims: 26
Claim 1
Independent Claim
197 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting the contour line of an object image comprises the steps of: i) setting a view window having a predetermined size smaller than the image, in a portion of the image having the candidate for a predetermined object image; ii) enlarging the size of the view window so that the entire candidate for the predetermined object image is within the view window, if the size of the predetermined object image is larger than the view window; iii) causing the center of the view window to travel to the center of the candidate for the predetermined object image; iv) fetching the candidate for a predetermined object image falling within said view window; v) normalizing the size and/or the shape of the fetched candidate for the predetermined object image based on the center point of said view window and a contour line of said candidate for the predetermined object image as references; and vi) thereafter extracting a contour line of the normalized candidate for the predetermined object image.
Claim 2
Independent Claim
197 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting the contour line of an object image comprises the steps of: i) setting a view window having a predetermined size smaller than the image, in a portion of the image having the candidate for a predetermined object image; ii) enlarging the size of the view window so that the entire candidate for the predetermined object image is within the view window, if the size of the predetermined object image is larger than the view window; iii) causing the center of the view window to travel to the center of the candidate for the predetermined object image; iv) fetching the candidate for a predetermined object image falling within said view window; v) normalizing the size and/or the shape of the fetched candidate for the predetermined object image based on the center point of said view window and a contour line of said candidate for the predetermined object image as references; and vi) thereafter extracting a contour line of the normalized candidate for the predetermined object image.
Claim 3
Dependent Claim
43 words
A method for extracting a contour line of an object image as defined in claim 1, wherein a neural network is employed in order to carry out at least a single step selected from the group consisting of steps i) ii) and iii).
Claim 4
Dependent Claim
31 words
A method for extracting a contour line of an object image as defined in claim 1, wherein said center point of said view window is centered on said predetermined object image.
Claim 5
Dependent Claim
34 words
A method for extracting a contour line of an object image as defined in claim 1, wherein said predetermined object image is mapped onto a complex-log plane prior to normalizing said predetermined object image.
Claim 6
Independent Claim
183 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines, which extend in a predetermined direction with respect to said viewpoint, from said image; and iii) causing lines, from among said lines detected in step ii), having a high level of continuity in approximately the same direction as said predetermined direction, to cooperate with one another thereby emphasizing said lines having a high level of continuity, and causing said lines having a high level of continuity and lines, from among said lines detected in step ii), having a low level of continuity in approximately the same direction as said predetermined direction to compete with each other, whereby said lines having a low level of continuity are erased, and said lines having a high level of continuity remain as the extracted contour line of said candidate for the predetermined object image.
Claim 7
Independent Claim
183 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines, which extend in a predetermined direction with respect to said viewpoint, from said image; and iii) causing lines, from among said lines detected in step ii), having a high level of continuity in approximately the same direction as said predetermined direction, to cooperate with one another thereby emphasizing said lines having a high level of continuity, and causing said lines having a high level of continuity and lines, from among said lines detected in step ii), having a low level of continuity in approximately the same direction as said predetermined direction to compete with each other, whereby said lines having a low level of continuity are erased, and said lines having a high level of continuity remain as the extracted contour line of said candidate for the predetermined object image.
Claim 8
Dependent Claim
55 words
A method for extracting a contour line of an object image as defined in claim 5, wherein step ii) of detecting said lines which extend in the predetermined direction is carried out after said image has been transformed with complex-log mapping into a complex-log mapped image by taking said viewpoint as a pole of mapping.
Claim 9
Dependent Claim
42 words
A method for extracting a contour line of an object image as defined in claim 5, wherein a neural network is employed in order to carry out at least a single step selected from the group consisting of steps ii) and iii).
Claim 10
Dependent Claim
178 words
A method for extracting a contour line of an object image as defined in claim 5, wherein step ii) of detecting lines employs a first plurality of synaptic weight patterns for detecting contour lines by performing a convolution operation on said image by using each of said first plurality of synaptic weight patterns for detecting contour lines, and wherein said step iii) of emphasizing and erasing said detected lines comprises the steps of: a) performing a convolution operation on each of groups of said lines detected in step ii), wherein said groups of lines are detected with said first plurality of synaptic weight patterns, by use of each of a second plurality of synaptic weight patterns for competition and cooperation, said second plurality of synaptic weight patterns having sizes corresponding to the sizes of said first plurality of synaptic weight patterns used for detecting contour lines, thereby emphasizing said groups of lines; and (b) causing said groups of lines to compete and cooperate with one another, thereby erasing said groups of lines exhibiting said low level of continuity.
Claim 11
Dependent Claim
34 words
A method for extracting a contour line of an object image as defined in claim 8, wherein said plurality of synaptic weight patterns have pattern sizes differing from one another in a step-wise manner.
Claim 12
Dependent Claim
125 words
A method for extracting a contour line of an object image as defined in claim 5 further comprises the step of: iv) causing lines, from among said lines detected in step ii), having a high intensity to cooperate with one another, thereby emphasizing said lines having a high intensity, causing said lines having a high intensity and lines having a low intensity from among said lines detected in step ii) to compete with each other, whereby said lines having a low intensity are erased and said lines having a high intensity remain, and combining said remaining lines having a high intensity with said remaining lines having a high level of continuity to form the extracted contour line of said candidate for the predetermined object image.
Claim 13
Independent Claim
109 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines, which extend in a radial direction with respect to said viewpoint, from said image; iii) detecting end points of said detected lines which extend in said radial direction; and iv) forming a contour line of said candidate for the predetermined object image, which line extends in an annular direction, in accordance with said detected end points.
Claim 14
Independent Claim
109 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines, which extend in a radial direction with respect to said viewpoint, from said image; iii) detecting end points of said detected lines which extend in said radial direction; and iv) forming a contour line of said candidate for the predetermined object image, which line extends in an annular direction, in accordance with said detected end points.
Claim 15
Dependent Claim
54 words
A method for extracting a contour line of an object image as defined in claim 11, wherein the detection of said lines, which extend in the radial direction, is carried out after said image has been transformed with complex-log mapping into a complex-log mapped image by taking said viewpoint as a pole of mapping.
Claim 16
Dependent Claim
43 words
A method for extracting a contour line of an object image as defined in claim 11, wherein a neural network is employed in order to carry out at least a single step selected from the group consisting of steps ii), iii) and iv).
Claim 17
Independent Claim
131 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines which extend in an annular direction and in a radial direction with respect to said viewpoint, from said image; iii) emphasizing only said detected lines which extend in the radial direction; iv) detecting end points of said detected lines which extend in the radial direction; and v) of said detected lines, emphasizing a contour line of said candidate for the predetermined object image, which contour line extends in the annular direction, in accordance with said detected end points.
Claim 18
Independent Claim
131 words
A method for extracting a contour line of an object image, wherein a contour line of a candidate for a predetermined object image is extracted from an image, the method for extracting a contour line of an object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines which extend in an annular direction and in a radial direction with respect to said viewpoint, from said image; iii) emphasizing only said detected lines which extend in the radial direction; iv) detecting end points of said detected lines which extend in the radial direction; and v) of said detected lines, emphasizing a contour line of said candidate for the predetermined object image, which contour line extends in the annular direction, in accordance with said detected end points.
Claim 19
Dependent Claim
59 words
A method for extracting a contour line of an object image as defined in claim 14, wherein the detection of said lines which extend in the radial direction and in the annular direction is carried out after said image has been transformed with complex-log mapping into a complex-log mapped image by taking said viewpoint as a pole of mapping.
Claim 20
Dependent Claim
44 words
A method for extracting a contour line of an object image as defined in claim 14, wherein a neural network is employed in order to carry out at least a single step selected from the group consisting of steps ii), iii), iv) and v).
Claim 21
Independent Claim
325 words
A method for extracting a contour line of an object image, wherein contour lines of a plurality of candidates for predetermined object images are extracted from an image, in which the plurality of candidates for predetermined object images are embedded such that at least portions of the plurality of candidates for predetermined object images may overlap one upon another, the method for extracting a contour line of the object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines which extend in a predetermined direction with respect to said viewpoint, from said image; iii) causing lines, from among said lines detected in step ii), having a high level of continuity in approximately the same direction as said predetermined direction, thereby emphasizing said lines having a high level of continuity, and causing said lines having a high level of continuity and lines having a low level of continuity in approximately the same direction as said predetermined direction to compete with each other, whereby said lines having a low level of continuity are erased, and said lines having a high level of continuity remain as an extracted contour line of a single candidate for the predetermined object image among the plurality of candidates for predetermined object images; iv) carrying out a process for preventing re-extraction on said extracted contour line of said single candidate for the predetermined object image; and v) performing step iii) to extract another contour line of another candidate for the predetermined object image, wherein said another contour line is different from said extracted contour line of said single candidate for the predetermined object image, the extraction being carried out by emphasizing lines which are different from said emphasized lines, and performing a process for preventing re-extraction of said extracted another contour line of said another candidate for the predetermined object image, unless contour lines of all of the plurality of said candidates for predetermined object images are extracted.
Claim 22
Independent Claim
325 words
A method for extracting a contour line of an object image, wherein contour lines of a plurality of candidates for predetermined object images are extracted from an image, in which the plurality of candidates for predetermined object images are embedded such that at least portions of the plurality of candidates for predetermined object images may overlap one upon another, the method for extracting a contour line of the object image comprises the steps of: i) setting a viewpoint in said image; ii) detecting lines which extend in a predetermined direction with respect to said viewpoint, from said image; iii) causing lines, from among said lines detected in step ii), having a high level of continuity in approximately the same direction as said predetermined direction, thereby emphasizing said lines having a high level of continuity, and causing said lines having a high level of continuity and lines having a low level of continuity in approximately the same direction as said predetermined direction to compete with each other, whereby said lines having a low level of continuity are erased, and said lines having a high level of continuity remain as an extracted contour line of a single candidate for the predetermined object image among the plurality of candidates for predetermined object images; iv) carrying out a process for preventing re-extraction on said extracted contour line of said single candidate for the predetermined object image; and v) performing step iii) to extract another contour line of another candidate for the predetermined object image, wherein said another contour line is different from said extracted contour line of said single candidate for the predetermined object image, the extraction being carried out by emphasizing lines which are different from said emphasized lines, and performing a process for preventing re-extraction of said extracted another contour line of said another candidate for the predetermined object image, unless contour lines of all of the plurality of said candidates for predetermined object images are extracted.
Claim 23
Dependent Claim
54 words
A method for extracting a contour line of an object image as defined in claim 17, wherein the detection of said lines which extend in the predetermined direction, is carried out after said image has been transformed with complex-log mapping into a complex-log mapped image by taking said viewpoint as a pole of mapping.
Claim 24
Dependent Claim
44 words
A method for extracting a contour line of an object image as defined in claim 17, wherein step iii) is performed by using a neural network, and step iv) for preventing re-extraction is carried out by introducing a fatigue term into the neural network.
Claim 25
Dependent Claim
137 words
A method for extracting a contour line of an object image as defined in claim 17 further comprising the step of: vi) causing lines, from among said lines detected in step ii), having a high 4 intensity to cooperate with one another, thereby emphasizing said lines having a high intensity, causing said lines having a high intensity and lines having a low intensity from among said lines detected in step ii) to compete with each other, whereby said lines having a low intensity are erased and said lines having a high intensity remain, and combining said remaining lines having a high intensity with said remaining lines having a high level of continuity to form the extracted contour line of a single candidate for the predetermined object image among the plurality of said candidates for predetermined object images.
Claim 26
Dependent Claim
45 words
A method for extracting a contour line of an object image as defined in claim 20, wherein a neural network is employed in order to carry out at least a single step selected from the group consisting of steps ii), iii), iv), v) and vi).