Method apparatus and system for determining a thermal signature
A method of determining a thermal signature from thermal data associated with a body section is disclosed. The method comprises: segmenting the thermal data into a plurality of segments, and calculating a set of locations defining a contour, each location being central with respect to picture-elements associated with one segment, thereby determining the thermal signature based on the contour.
1. Apparatus for identifying blood vessels in a thermal image of a section of a living body, the apparatus comprising:
a convolution unit, for convolving intensity data representing the thermal image with a predetermined vessel shapes filter thereby to provide filtered data;
a derivative calculator, for calculating, for each picture-element of said thermal image, calculating at least a first local derivative of said filtered data along a first dimension and a second local derivative of said filtered data along a second dimension, and selecting a highest of said first and said second local derivatives, thereby providing derivative data;
a local intensity extrema searcher, for searching in said derivative data for local intensity extrema; and
an interpolator, for applying an interpolation procedure for generating contours between at least a portion of said local intensity extrema, and identifying said contours as blood vessels.
2. The apparatus of claim 1 , further comprising an intensity data inverter for inverting said intensity data by linear transformation.
3. The apparatus of claim 1 , further comprising a masking unit for masking said intensity data so as to exclude at least a portion of said intensity data, said portion corresponding to picture-elements not belonging to blood vessels.
4. The apparatus of claim 3 , wherein said masking unit is operable to calculate at least one of: local minimum, local maximum and local average for each picture-element of the thermal image.
5. The apparatus of claim 1 , further comprising a normalization unit for normalizing said intensity data.
6. The apparatus of claim 1 , further comprising a noise reduction unit which employs a noise removal procedure to exclude isolated local intensity extrema.
7. A method of identifying blood vessels in a thermal image of a section of a living body, the method comprising:
convolving intensity data representing the thermal image with a predetermined vessel shapes filter thereby providing filtered data;
for each picture-element of said thermal image, calculating at least a first local derivative of said filtered data along a first dimension and a second local derivative of said filtered data along a second dimension, and selecting a highest of said first and said second local derivatives, thereby providing derivative data;
searching in said derivative data for local intensity extrema; and
applying an interpolation procedure for generating contours between at least a portion of said local intensity extrema, and identifying said contours as blood vessels.
8. The method of claim 7 , further comprising, prior to said convolution, inverting said intensity data by linear transformation.
9. The method of claim 7 , further comprising masking said intensity data so as to exclude at least a portion of said intensity data, said portion corresponding to picture-elements not belonging to blood vessels.
10. The method of claim 9 , wherein said masking comprises calculating at least one of: local minimum, local maximum and local average for each picture-element of the thermal image.
11. The method of claim 7 , further comprising normalizing said intensity data.
12. The method of claim 7 , further comprising, subsequently to said search for local intensity extrema, employing a noise reduction procedure for excluding isolated local intensity extrema.
13. The method of claim 7 , further comprising generating a blood vessel map based on said identified blood vessels.
14. A method of recognizing an individual based on a body section of the individual, comprising:
identifying blood vessels in a thermal image of the body section using the method of claim 7 , so as to generate a blood vessel map;
searching a searchable database of blood vessel maps for a map entry which is similar to said blood vessel map; and
identifying the individual based on said map entry.
15. A method of estimating characteristic heat conduction of a section of a living body, comprising:
obtaining a series of thermospatial representations describing the section of the living body while having at least two different shapes;
for each shape of said at least two different shapes, executing the method of claim 7 for identifying at least one blood vessel in thermal data of a respective thermospatial representation and calculating a depth of said at least one blood vessel;
determining thermal stabilization period for said at least one blood vessel; and
determining the characteristic heat conduction based at least in part on said thermal stabilization period.