Method, apparatus, computer program, computer system and computer-readable storage for representing and searching for an object in an image
A method of representing an object appearing in a still or video image, by processing signals corresponding to the image, comprises deriving a curvature scale space (CSS) representation of the object outline by smoothing the object outline, deriving at least one additional parameter reflecting the shape or mass distribution of a smoothed version of the original curve, and associating the CSS representation and the additional parameter as a shape descriptor of the object.
1. A method of creating a shape descriptor representing an object appearing in an image or sequence of images, by processing signals corresponding to the image or images using at least one processor, the method comprising:
deriving, using a processor, a curvature scale space (CSS) representation of an original outline of the object by smoothing the outline;
deriving, using a processor, an additional parameter representing the shape or mass distribution of a smoothed version of the original outline of the object, said smoothed version of the original outline of the object corresponding to a peak in a CSS image of the CSS representation; and
associating, using a processor, the CSS representation and the additional parameter to form the shape descriptor of the object.
2. A method as claimed in claim 1 wherein the smoothed version of the original object outline corresponds to the highest peak in the CSS image.
3. An apparatus comprising:
an image database storing images or sequences of images;
a descriptor database storing descriptors of objects or parts of objects appearing in images stored in the image database; and
a processor configured to perform the method as claimed in claim 1 .
4. A computer-readable storage medium storing computer-executable process steps for implementing a method as claimed in claim 1 .
5. A computer system comprising:
an image database storing images or sequences of images;
a descriptor database storing descriptors of objects or parts of objects appearing in images stored in the image database; and
a control unit programmed to perform the method as claimed in claim 1 .
6. A method of deriving a shape descriptor representing a plurality of objects appearing in an image or sequence of images, by processing signals corresponding to the image or images using at least one processor, the method comprising:
deriving, using a processor, a curvature scale space (CSS) representation for each of the plurality of objects;
for each object, determining, using a processor, if there is at least one peak in an CSS image of the CSS representation, and, if there is at least one peak in the CSS image, then deriving a shape descriptor using a method as claimed in claim 1 ; and
if there are no peaks in the CSS image, then deriving, using a processor, a shape descriptor including at least an additional parameter representing the shape or mass distribution of a smoothed version of an original object outline.