Lossless variable-bit signature compression
View Patent ↗Apparatus and methods for compressing data points. The invention compresses the representation of a sequence of points in a space by, for example, dividing the sequence of points into segments of successive points and then compressing each of the segments irrespective of the compression applied to the other segments. The dividing may include dividing the points sequence into segments of S successive points. The value of S may be the value of S generating the smallest of the multiple compressions. The compression may include compressing each of the segments of S successive, i-bit points into segments of j-bit points, where j<=i,j may vary from segment to segment and, for any given segment, j is the minimum number of bits necessary to represent the data in that given segment.
1. A method for compressing a representation of a sequence of points in a space, the method comprising:
dividing the sequence of points into segments of successive points;
determining a compression size for each of the segments, wherein the compression size varies based on a number of bits needed to represent relative distances between the points of each segment; and
compressing each of the segments into the compression size for each segment.
2. The method of claim 1 , wherein the step of dividing comprises dividing a sequence of points into segments of S successive, i-bit points.
3. The method of claim 2 , wherein before the step of dividing, the following step is performed:
determining the value of S.
4. The method of claim 3 , wherein the step of determining the value of S comprises generating multiple compressions of the sequence, each of the multiple compressions at a different value of S.
5. The method of claim 3 , wherein the step of determining the value of S comprises generating a compression of the sequence for each value of S from a minimum to a maximum.
6. The method of claim 3 , wherein the step of determining the value of S comprises generating a compression of the sequence for each value of S from a minimum of two (2) to a maximum equal to the number of points in the sequence.
7. The method of claim 3 , wherein the step of determining the value of S comprises
generating multiple compressions of the sequence, each of the multiple compressions at a different value of S; and
determining the value of S to be the value of S generating the smallest of the multiple compressions.
8. The method of claim 1 , wherein the step of compressing comprises
compressing each of the segments of S successive, i-bit points into segments of j-bit points, where j<=i.
9. The method of claim 8 , wherein the value of j may vary from segment to segment.
10. The method of claim 8 , wherein, for any given segment, j is the minimum number of bits necessary to represent the data in that given segment.
11. The method of claim 1 , wherein the step of compressing comprises
determining the largest coordinate in any dimension of any point in a segment;
setting j for the segment to the ceiling of the base- 2 log of that largest coordinate; and
truncating from points of the segment most significant bits exceeding j bits.
12. The method of claim 1 , wherein the sequence of points is an electronic signature.
13. The method of claim 1 , wherein the step of compressing comprises
compressing each of the segments without losing any of the data in any of the segments.
14. The method of claim 1 , wherein the step of compressing comprises
compressing each of the segments, losing data as directed by an invoking user.
15. The method of claim 1 , wherein before the step of dividing the following Step is performed:
convening DrawTo data to relative-movement data.
16. A computer readable medium wherein is located a computer program for compressing a representation of a sequence of points in a space by:
dividing the sequence of points into segments of successive points;
determining a compression size for each of the segments, wherein the compression size varies based on a number of bits needed to represent relative distances between the points of each segment; and
compressing each of the segments into the compression size for each segment.
17. A compressor for compressing a representation of a sequence of points in a space, comprising:
a computer readabl& medium wherein is located a computer program for compressing the representation of the sequence of points in the space by:
dividing the sequence of points into segments of successive points;
determining a compression size for each of the segments, wherein the compression size varies based on a number of bits needed to represent relative distances between the points of each segment; and
compressing each of the segments into the compression size for each seament;
a CPU for executing the computer program in the data store; and
a link, communicatively coupling the data store and the CPU.