1. A cellulose ester film having a front retardation Re of from 20 nm to 100 nm relative to light having a wavelength of 590 nm at 25\xb0 C. and 60% RH; a retardation Rth in a film thickness direction of from 60 nm to 400 nm relative to light having a wavelength of 590 nm at 25\xb0 C. and 60% RH; and a shrinkage factor in a slow axis direction of less than 1% in the range of from a glass transition temperature to a temperature of 40\xb0 C. higher than a glass transition temperature.
2. The cellulose ester film according to claim 1, wherein the cellulose ester film has an acyl group having from 2 to 4 carbon atoms as a substituent and, when a degree of substitution of an acetyl group is defined as DSA and a degree of substitution of a propionyl group or a butyryl group is defined as DSB, is satisfied with a relationship of {2.0\u2266(DSA+DSB)\u22662.9}.
3. The cellulose ester film according to claim 2, wherein the cellulose ester film has an acyl group having from 2 to 4 carbon atoms as a substituent and, when a degree of substitution of an acetyl group is defined as DSA and a degree of substitution of a propionyl group or a butyrl group is defined as DSB, is satisfied with relationships of {2.0\u2266(DSA+DSB)\u22662.9} and {0.4\u2266DSB\u22661.0}.
4. The cellulose ester film according to claim 1, which comprises an aliphatic polyhydric alcohol ester formed from an aliphatic polyhydric alcohol and one or more monocarboxylic acid.
5. The cellulose ester film according to claim 1, which comprises at least one of a retardation developing agent composed of a rod-like or disc-like compound.
6. The cellulose ester film according to claim 1, wherein the film has a thickness of from 20 to 100 \u03bcm.
7. A polarizing plate comprising the cellulose ester film according to claim 1 stuck onto at least one surface of a polarizer or a polarizing plate.
8. A liquid crystal display comprising the cellulose ester film according to claim 1.
9. A process for producing a cellulose ester film comprising: a fabrication step of casting a solution having a cellulose ester dissolved in an organic solvent on a support and evaporating the solvent to form a cellulose ester film; a stretching step of subsequently stretching the film; and a drying step of drying the resulting film, which further comprises a step of, after completion of the drying step, performing a heat treatment at a temperature of from 150 to 200\xb0 C. for one minute or more.
10. The process for producing a cellulose ester film according to claim 9, wherein a stretch ratio of the film in the stretching step is from 1.2 to 2.0 times.
11. The process for producing a cellulose ester film according to claim 9, wherein in the heat treatment step, the film is stretched from 0.9 to 1.5 times in a width direction and a traveling direction of the film, respectively.
12. A liquid crystal display comprising the polarizing plate according to claim 7.
The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.
What is claimed is:
1. An apparatus of generating motion data, comprising:
a storage unit for storing plural time-series motion data; and
a motion connecting unit for reading out plural time-series motion data from said storage unit, connecting the plural time-series motion data to generate a series of time-series motion data, and storing the series of time-series motion data in said storage unit.
2. An apparatus of generating motion data, comprising:
a storage unit for storing two time-series motion data Lo be connected, the two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively; and
a motion connecting unit for estimating motion data in future or past as much as a connection time, in terms of time, for either the front time-series motion data or the back time-series motion data, to generate estimated-extended motion data, and synthesizing connecting motion data based on the estimated-extended motion data and motion data in future or past as much as the connection time, in terms of time, for either the front time-series motion data or the back time-series motion data.
3. The apparatus of claim 2 wherein said motion connecting unit, two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively, estimates motion data as much as a connection time from the last time of the front time-series motion data to generate estimated-extended motion data, and generates connecting motion data based on the back time-series motion data from the initial time to the connection time after and the estimated-extended motion data.
4. The apparatus of claim 2 wherein said motion connecting unit, two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively, estimates motion data as much as a connection time in the past direction from the initial time of the back time-series motion data, in terms of time, to generate estimated-extended motion data, and generates connecting motion data based on the front time-series motion data from the last time to the connection time before and the estimated-extended motion data.
5. The apparatus of claim 2 wherein said motion connecting unit, according to a connection time distribution rate , (a) estimates motion data as much as (a connection time) from the last time of the front time-series motion data to generate estimated backward-extended motion data, (b) estimates motion data in the initial time of the back time-series motion data, in terms of time, to generate estimated forward-extended motion data, (c) generates a first half of connecting motion data based on the front time-series motion data from the last time to (1-)(the connection time) before and the estimated forward-extended motion data, and (d) generates a second half of connecting motion data based on the back time-series motion data from the initial time to (the connection time) after and the estimated backward-extended motion data.
6. The apparatus of any of claim 3, claim 4, and claim 5 wherein said open periodic motion generating unit uses a function that (a) is 0 at the initial time of the connection time, and 1 at the last time of the connection time, (b) increases monotonically, (c) is differentiable, and (d) is rotationally symmetrical by 180 degrees about the middle time of the connection time, within the connection time.
7. The apparatus of any of claim 3, claim 4, and claim 5 wherein said open periodic motion generating unit uses a function that (a) is 1 at the initial time of the connection time, and 0 at the last time of the connection time, (b) decreases monotonically, (c) is differentiable, and (d) is rotationally symmetrical by 180 degrees about the middle time of the connection time, within the connection time.
8. The apparatus of claim 3 wherein said motion connecting unit estimates that the front time-series motion data at the last time continues as much as a connection time from the last time, and generates estimated-extended motion data.
9. The apparatus of claim 4 wherein said motion connecting unit estimates that the back time-series motion data at the initial time has continued as much as a connection time in the past direction from the initial time, in terms of time, and generates estimated-extended motion data.
10. The apparatus of claim 3 wherein said motion connecting unit estimates that motion data smoothly shifts to the front time-series motion data at the last time a connection time after the last time, and generates estimated-extended motion data.
11. The apparatus of claim 4 wherein said motion connecting unit estimates that the motion state of the back time-series motion data a connection time before the initial time of the back time-series motion data smoothly shifts to the back time-series motion data with the motion state of the back time-series motion data at the initial time, and generates estimated-extended motion data.
12. The apparatus of claim 3 wherein said motion connecting unit estimates that the motion state of the front time-series motion data a connection time after the last time of the front time-series motion data is the motion state of the back time-series motion data the connection time after the initial time of the back time-series motion data, and generates estimated-extended motion data.
13. The apparatus of claim 4 wherein said motion connecting unit estimates that the motion state of the back time-series motion data a connection time before the initial time of the back time-series motion data smoothly shifts to the motion state of the front time-series motion data the connection time before the last time of the front time-series motion data, and generates estimated-extended motion data.
14. The apparatus of claim 5 wherein said motion connecting unit calculates a connection time distribution rate in connection, for time-series motion data holding a connection time distribution rate as additional information, based on d connection time distribution rate as additional information for the front time-series motion data and a connection time distribution rate as additional information for the back time-series motion data, and generates an estimated forward-extended motion and an estimated backward-extended motion according to the connection time distribution rate.
15. The apparatus of claim 5 wherein said motion connecting unit calculates a connection time distribution rate in connection, for time-series motion data holding a connection time distribution rate as additional information, based on a connection time distribution rate 1 as additional information for the front time-series motion data and a connection time distribution rate 2 as additional information for the back time-series motion data, by (1 2)2, and generates an estimated forward-extended motion and an estimated backward-extended motion according to the connection time distribution rate.
16. The apparatus of any of claim 3, claim 4, and claim 5 wherein for time-series motion data holding additional information about numerically expressed methods of generating estimated-extended motions, said motion connecting unit generates an estimated-extended motion, an estimated backward-extended motion, and an estimated forward-extended motion, according to the additional information.
17. The apparatus of claim 11 wherein said motion connecting unit, for time-series motion data holding a value numerically expressing the importance of the time-series motion data as additional information, based on the value of the importance of the additional information for the front time-series motion data and the value of tile importance of the additional information for the back time-series motion data, determines a method of generating an estimated backward-extended motion and an estimated forward-extended motion, calculates a connection time distribution rate, and generates an estimated forward-extended motion and an estimated backward-extended motion.
18. The apparatus of claim 5 wherein said motion connecting unit, for time-series motion data holding a value numerically expressing the importance of the time-series motion data as additional information, based on the value 1 of the importance of the additional information for the front time-series motion data and the value 2 of the importance of the additional information for the back time-series motion data (01, 21), determines a method of generating an estimated backward-extended motion and an estimated forward-extended motion, calculates a connection time distribution rate by (121)2, and generates an estimated forward-extended motion and an estimated backward-extended motion.
19. An apparatus of generating closed-loop periodic motion data, comprising:
a storage unit;
an inputoutput unit for (a) receiving time-series motion data including the position data of a whole multi-joint rigid body object at each time, and the posture angle data, joint slide vector data, and joint angle data of the multi-joint rigid body object at each time, (h) storing the received time-series motion data in said storage unit, and (c) reading out closed-loop periodic motion data generated from the time-series motion data, stored in said storage unit;
an open periodic motion generating unit for (a) reading out time-series motion data stored in said storage unit, (b) generating open periodic time-series motion data by making the local parts of a multi-joint rigid body object periodic, and (c) storing the open periodic time-series motion data in said storage unit; and
a closed-loop motion generating unit for (a) reading out the position data of the whole multi-joint rigid body object among the open periodic time-series motion data, (b) generating closed-loop motion data by making the position data of the whole multi-joint rigid body object closed-loop, i.e., the movement of the whole multi-joint rigid body object being made closed-loop, (c) replacing the open periodic time-series motion data stored in said storage unit with the closed-loop motion data, and (d) storing the closed-loop motion data in said storage unit.
20. The apparatus of claim 19 wherein said apparatus further includes a motion connecting unit for (a) reading out plural time-series motion data stored said storage unit via said data bus, (b) connecting the plural time-series motion data to generate a series of time-series motion data, and (c) storing the series of time-series motion data in said storage unit via said data bus.
21. The apparatus of any of claim 19 and claim 20 wherein said open periodic motion generating unit (a) estimates motion data as much as a synthesis time from the last time of time-series motion data to generate estimated-extended motion data, (b) combines the time-series motion data from the initial time to the synthesis time after and the estimated-extended motion data to generate synthesized motion data, (c) replaces the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the initial time to the synthesis time after with those of the synthesized motion data, and (d) calculates the amount of parallel translation for the position data of the whole multi-joint rigid body object, performs the parallel translation to the position data of the whole multi-joint rigid body object, and replaces the position data of the time-series motion data from the initial time to the synthesis time after with the resulting position data.
22. The apparatus of any of claim 19 and claim 20 wherein said open periodic motion generating unit (a) estimates motion data as much as a synthesis time in the past direction from the initial time of time-series motion data, in terms of time, to generate estimated-extended motion data, (b) combines the time-series motion data from the last time to the synthesis time before and the estimated-extended motion data to generate synthesized motion data, (c) replaces the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to the synthesis time before with those of the synthesized motion data, and (d) calculates the amount of parallel translation for the position data of the whole multi-joint rigid body object, performs the parallel translation to the position data of the whole multi-joint rigid body object, and replaces the position data of the time-series motion data from the last time to the synthesis time before with the resulting position data.
23. The apparatus of any of claim 19 and claim 20 wherein said open periodic motion generating unit (a) estimates notion data as much as (a synthesis time) from the last time of time-series motion data, being a synthesis time distribution rate (01), to generate estimated backward-extended motion data, (b) estimates motion data as much as (1)(the synthesis time) in the past direction from the initial time of time-series motion data, in terms of time, to generate estimated forward-extended motion data, (c) combines the time-series motion data from the initial time to (the synthesis time) after and the estimated backward-extended motion data to generate front synthesized motion data, (d) combines the time-series motion data from the last time to (1)(the synthesis time) before and the estimated forward-extended motion data to generate back synthesized motion data, (e) replaces the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to (the synthesis time) after with those of the front synthesized motion data, (f) calculates the amount of parallel translation for the position data of the whole multi-joint rigid body object from the last time to (the synthesis time) after, performs the parallel translation to the position data of the whole multi-joint rigid body object from the last time to (the synthesis time) after, and replaces the position data of the time-series motion data from the last time to (the synthesis time) after with the resulting position data, (g) replaces the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to (1)(the synthesis time) before with those of the back synthesized motion data, and (h) calculates the amount of parallel translation for the position data of the whole multi-joint rigid body object from the last time to (1)(the synthesis time) before, performs the parallel translation to the position data of the whole multi-joint rigid body object from the last time to (1)(the synthesis time) before, and replaces the position data of the time-series motion data from the last time to (1)(the synthesis time) before with the resulting position data.
24. The apparatus of any of claim 19 and claim 20 wherein said closed-loop motion generating unit performs parallel translation, using a monotonically increasing function, to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, in a way to match the position data of the whole multi-joint rigid body object at the last time to that at the initial time.
25. The apparatus of any of claim 19 and claim 20 wherein said closed-loop motion generating unit performs parallel translation to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, by applying to the position data of the whole multi-joint rigid body object at a time t within the effect time, an amount of parallel translation obtained by multiplying the difference resulting from subtracting the position data of the whole multi-joint rigid body object at the initial time from that at the last time, with the difference between t and the initial time of the effect time, divided by the effect time.
26. The apparatus of any of claim 19 and claim 20 wherein said closed-loop motion generating unit performs parallel translation to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, by applying to the position data of the whole multi-joint rigid body object at a time t within the effect time, an amount of parallel translation obtained by multiplying the difference resulting from subtracting the position data of the whole multi-joint rigid body object at the initial time from that at the last time, with the difference between t and the initial time of the effect time, divided by the effect time, and raising the result to the power of (1).
27. The apparatus of any of claim 21, claim 22, and claim 23 wherein said open periodic motion generating unit uses a function that (a) is 1 at the initial time of the synthesis time, and 0 at the last time of the synthesis time, (b) decreases monotonically, (c) is differentiable, and (d) is rotationally symmetrical by 180 degrees about the middle time of the synthesis time, within the synthesis time.
28. The apparatus of any of claim 21, claim 22, and claim 23 wherein said open periodic motion generating unit uses a function that (a) is 0 at the initial time of the synthesis time, and 1 at the last time of the synthesis time, (b) increases monotonically, (c) is differentiable, and (d) is rotationally symmetrical by 180 degrees about the middle time of the synthesis time, within the synthesis time.
29. The apparatus of any of claim 19 and claim 20 wherein said inputoutput unit receives time-series motion data compressed in advance, and said apparatus further includes a decompression unit for decompressing the compressed time-series motion data, and storing the decompressed time-series motion data in said storage unit.
30. A method of generating motion data, comprising:
storing plural time-series motion data; and
connecting motions by reading out plural time-series motion data from said storage unit, connecting the plural time-series motion data to generate a series of time-series motion data, and storing the series of time-series motion data in said storage unit.
31. A method of generating motion data, comprising:
storing two time-series motion data to be connected, the two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively; and
connecting motions by estimating motion data in future or past as much as a connection time, in terms of time, for either the front time-series motion data or the back time-series motion data, to generate estimated-extended motion data, and synthesizing connecting motion data based on the estimated-extended motion data and motion data in future or past as much as the connection time, in terms of time, for either the front time-series motion data or the back time-series motion data.
32. The method of claim 31 wherein said connecting motions, two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively, includes estimating motion data as much as a connection time from the last time of the front time-series motion data to generate estimated-extended motion data, and generating connecting motion data based on the back time-series motion data from the initial time to the connection time after and the estimated-extended motion data.
33. The method of claim 31 wherein said connecting motions, two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively, includes estimating motion data as much as a connection time in the past direction from the initial time of the back time-series motion data, in terms of time, to generate estimated-extended motion data, and generating connecting motion data based on the front time-series motion data from the last time to the connection time before and the estimated-extended motion data.
34. The method of claim 31 wherein said connecting motions includes, according to a connection time distribution rate , (a) estimating motion data as much as (a connection time) from the last time of the front time-series motion data to generate estimated backward-extended motion data, (b) estimating motion data in the past direction as much as (1-)(the connection time) from the initial time of the back time-series motion data, in terms of time, to generate estimated forward-extended motion data, (c) generating a first half of connecting motion data based on the front time-series motion data from the last time to (1-)(the connection time) before and the estimated forward-extended motion data, and (d) generating a second half of connecting motion data based on the back time-series motion data from the initial time to (the connection time) after and the estimated backward-extended motion data.
35. The method of any of claim 32, claim 33, and claim 34 wherein said connecting motions uses a function that (a) is 0 at the initial time of the connection time, and 1 at the last time of the connection time, (b) increases monotonically, (c) is differentiable, arid (d) is rotationally symmetrical by 180 degrees about the middle time of the connection time, within the connection time.
36. The method of any of claim 32, claim 33, and claim 34 wherein said connecting motions uses a function that (a) is 1 at the initial time of the connection time, and 0 at the last time of the connection time, (b) decreases monotonically, (c) is differentiable, and (d) is rotationally symmetrical by 180 degrees about the middle time of the connection time, within the connection time.
37. The method of claim 32 wherein said connecting motions includes estimating that the front time-series motion data at the last time continues as much as a connection time from the last time, and generating estimated-extended motion data.
38. The method of claim 33 wherein said connecting motions includes estimating that the back time-series motion data at the initial time has continued as much as a connection time in the past direction from the initial time, in terms of time, and generating estimated-extended motion data.
39. The method of claim 32 wherein said connecting motions estimating that motion data smoothly shifts to the front time-series motion data at the last time a connection time after the last time, and generating estimated-extended motion data.
40. The method of claim 33 wherein said connecting motions includes estimating that the motion state of the back time-series motion data a connection time before the initial time of the back time-series motion data smoothly shifts to the back time-series motion data with the motion state of the back time-series motion data at the initial time, and generating estimated-extended motion data.
41. The method of claim 32 wherein said connecting motions includes estimating that the motion state of the front time-series motion data a connection time after the last time of the front time-series motion data is the motion state of the back time-series motion data the connection time after the initial time of the back time-series motion data, and generating estimated-extended motion data.
42. The method of claim 33 wherein said connecting motions includes estimating that the motion state of the back time-series motion data a connection time before the initial time of the back time-series motion data smoothly shifts to the motion state of the front time-series motion data the connection time before the last time of the front time-series motion data, and generating estimated-extended motion data.
43. The method of claim 34 wherein said connecting motions includes calculating a connection time distribution rate in connection, for time-series motion data holding a connection time distribution rate as additional information, based on a connection time distribution rate as additional information for the front time-series motion data and a connection time distribution rate as additional information for the back time-series motion data, and generating an estimated forward-extended motion and an estimated backward-extended motion according to the connection time distribution rate.
44. The method of claim 34 wherein said connecting motions includes calculating a connection time distribution rate in connection, for time-series motion data holding a connection time distribution rate 1, as additional information, based on a connection time distribution rate 1 as additional information for the front time-series motion data and a connection time distribution rate 2 as additional information for the back time-series motion data, by (12)2, and generating an estimated forward-extended motion and an estimated backward-extended motion according to the connection time distribution rate.
45. The method of any of claim 32, claim 33, and claim 34 wherein for time-series motion data holding additional information about numerically expressed methods of generating estimated-extended motions, said connecting motions includes generating an estimated-extended motion, an estimated backward-extended motion, and an estimated forward-extended motion, according to the additional information.
46. The method of claim 34 wherein said connecting motions includes, for time-series motion data holding a value numerically expressing the importance of the time-series motion data as additional information, based on the value of the importance of the additional information for the front time-series motion data and the value of the importance of the additional information for the back time-series motion data, determining a method of generating an estimated backward-extended motion and an estimated forward-extended motion, calculating a connection time distribution rate, and generating an estimated forward-extended motion and an estimated backward-extended motion.
47. The method of claim 34 wherein said connecting motions includes, for time-series motion data holding a value numerically expressing the importance of the time-series motion data as additional information, based on the value 1 of the importance of the additional information for the front time-series motion data and the value 2 of the importance of the additional information for the back time-series motion data (01, 21), determining a method of generating an estimated backward-extended motion and an estimated forward-extended motion, calculating a connection time distribution rate by (121)2, and generating an estimated forward-extended motion and an estimated backward-extended motion.
48. A method of generating closed-loop periodic motion data, comprising:
administrating inputs and outputs by (a) receiving time-series motion data including the position data of a whole multi-joint rigid body object at each time, and the posture angle data, joint slide vector data, and the joint angle data of the multi-joint rigid body object at each time, (b) storing the received time-series motion data in a storage unit, and (c) reading out closed-loop periodic motion data generated from the time-series motion data, stored in said storage unit;
generating an open periodic motion by (a) reading out time-series motion data stored in said storage unit, (b) generating open periodic time-series motion data by making the local parts of a multi-joint rigid body object periodic, and (c) storing the open periodic time-series motion data in said storage unit; and
generating a closed-loop motion generating by (a) reading out the position data of the whole multi-joint rigid body object among the open periodic time-series motion data, (b) generating closed-loop motion data by making the position data of the whole multi-joint rigid body object closed-loop, i.e., the movement of the whole multi-joint rigid body object being made closed-loop, (c) replacing the open periodic time-series motion data stored in said storage unit with the closed-loop motion data, and (d) storing the closed-loop motion data in said storage unit.
49. The method of claim 48 wherein said method further includes connecting motions by (a) reading out plural time-series motion data stored said storage unit via said data bus, (b) connecting the plural time-series motion data to generate a series of time-series motion data, and (c) storing the series of time-series motion data in said storage unit via said data bus.
50. The method of any of claim 48 and claim 49 wherein said generating an open periodic motion includes (a) estimating motion data as much as a synthesis time from the last time of time-series motion data to generate estimated-extended motion data, (b) combining the time-series motion data from the initial time to the synthesis time after and the estimated-extended motion data to generate synthesized motion data, (c) replacing the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the initial time to the synthesis time after with those of the synthesized motion data, and (d) calculating the amount of parallel translation for the position data of the whole multi-joint rigid body object, performing the parallel translation to the position data of the whole multi-joint rigid body object, and replacing the position data of the time-series motion data from the initial time to the synthesis time after with the resulting position data.
51. The method of any of claim 48 and claim 49 wherein said generating an open periodic motion includes (a) estimating motion data as much as a synthesis time in the past direction from the initial time of time-series motion data, in terms of time, to generate estimated-extended motion data, (b) combining the time-series motion data from the last time to the synthesis time before and the estimated-extended motion data to generate synthesized motion data, (c) replacing the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to the synthesis time before with those of the synthesized motion data, and (d) calculating the amount of parallel translation for the position data of the whole multi-joint rigid body object, performing the parallel translation to the position data of the whole multi-joint rigid body object, and replacing the position data of the time-series motion data from the last time to the synthesis time before with the resulting position data.
52. The method of any of claim 48 and claim 49 wherein said generating open periodic motion includes (a) estimating motion data as much as (a synthesis time) from the last time of time-series motion data, a being a synthesis time distribution rate (01), to generate estimated backward-extended motion data, (b) estimating motion data as much as (1)(the synthesis time) in the past direction from the initial time of time-series motion data, in terms of time, to generate estimated forward-extended motion data, (c) combining the time-series motion data from the initial time to (the synthesis time) after and the estimated backward-extended motion data to generate front synthesized motion data, (d) combining the time-series motion data from the last time to (1)(the synthesis time) before and the estimated forward-extended motion data to generate back synthesized motion data, (e) replacing the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to (the synthesis time) after with those of the front synthesized motion data, (f) calculating the amount of parallel translation for the position data of the whole multi-joint rigid body object from the last time to (the synthesis time) after, performing the parallel translation to the position data of the whole multi-joint rigid body object from the last time to (the synthesis time after, and replacing the position data of the time-series motion data from the last time to (the synthesis time) after with the resulting position data, (q) replacing the posture angle data, joint slide vector data, and joint angle data of the time-series motion data from the last time to (1)(the synthesis time) before with those of the back synthesized motion data, and (h) calculating the amount of parallel translation for the position data of the whole multi-joint rigid body object from the last time to (1)(the synthesis time) before, performing the parallel translation to the position data of the whole multi-joint rigid body object from the last time to (1)(the synthesis time) before, and replacing the position data of the time-series motion data from the last time to (1)(the synthesis time) before with the resulting position data.
53. The method of any of claim 48 and claim 49 wherein in said generating a closed-loop motion, parallel translation is performed, using a monotonically increasing function, to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, in a way to match the position data of the whole multi-joint rigid body object at the last time to that at the initial time.
54. The method of any of claim 48 and claim 49 wherein in said generating a closed-loop motion, parallel translation is performed to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, by applying to the position data of the whole multi-joint rigid body object at a time t within the effect time, an amount of parallel translation obtained by multiplying the difference resulting from subtracting the position data of the whole multi-joint rigid body object at the initial time from that at the last time, with the difference between t and the initial time of the effect time, divided by the effect time.
55. The method of any of claim 48 and claim 49 wherein in said generating a closed-loop motion, parallel translation is performed to the position data of the whole multi-joint rigid body object from the last time to an effect time before, among the time-series motion data, by applying to the position data of the whole multi-joint rigid body object at a time t within the effect time, an amount of parallel translation obtained by multiplying the difference resulting from subtracting the position data of the whole multi-joint rigid body object at the initial time from that at the last time, with the difference between t and the initial time of the effect time, divided by the effect time, and raising the result to the power of (1).
56. The method of any of claim 50, claim 51, and claim 52 wherein in said generating an open periodic motion, used is a function that (a) is 1 at the initial time of the synthesis time, and 0 at the last time of the synthesis time, (b) decreases monotonically, (c) is differentiable, and (d) is rotationally symmetrical about the middle time of the synthesis time, within the synthesis time.
57. The method of any of claim 50, claim 51, and claim 52 wherein in said generating an open periodic motion, used is a function that (a) is 0 at the initial time of the synthesis time, and 1 at the last time of the synthesis time, (b) increases monotonically, (c) is differentiable, and (d) is rotationally symmetrical about the middle time of the synthesis time, within the synthesis time.
58. The method of any of claim 57 and claim 58 wherein said administrating inputs and outputs includes receiving time-series motion data compressed in advance, and said method further includes decompressing the compressed time-series motion data, and storing the decompressed time-series motion data in said storage unit.
59. A storage medium of storing a program of generating motion data, said program comprising:
storing plural time-series motion data; and
connecting motions by reading out plural time-series motion data from said storage unit, connecting the plural time-series motion data to generate a series of time-series motion data, and storing the series of time-series motion data in said storage unit.
60. A storage medium of storing a program of generating motion data, said program comprising:
storing two time-series motion data to be connected, the two time-series motion data being referred to as front time-series motion data and back time-series motion data, respectively; and
connecting motions by estimating motion data in future or past as much as a connection time, in terms of time, for either the front time-series motion data or the back time-series motion data, to generate estimated-extended motion data, and synthesizing connecting motion data based on the estimated-extended motion data and motion data in future or past as much as the connection time, in terms of time, for either the front time-series motion data or the back time-series motion data.
61. A storage medium of storing a program of generating closed-loop periodic motion data, said program comprising:
administrating inputs and outputs by (a) receiving time-series motion data including the position data of a whole multi-joint rigid body object at each time, and the posture angle data, joint slide vector data, and the joint angle data of the multi-joint rigid body object at each time, (b) storing the received time-series motion data in a storage unit via a data bus, and (c) reading out closed-loop periodic motion data generated from the time-series motion data, stored in said storage unit via said data bus;
generating an open periodic motion by (a) reading out time-series motion data stored in said storage unit via said data bus, (b) generating open periodic time-series motion data by making the local parts of a multi-joint rigid body object periodic, and (c) storing the open periodic time-series motion data in said storage unit via said data bus; and
generating a closed-loop motion generating by (a) reading out the position data of the whole multi-joint rigid body object among the open periodic time-series motion data, (b) generating closed-loop motion data by making the position data of the whole multi-joint rigid body object closed-loop, i.e., the movement of the whole multi-joint rigid body object being made closed-loop, (c) replacing the open periodic time-series motion data stored in said storage unit with the closed-loop motion data, and (d) storing the closed-loop motion data in said storage unit.
62. The storage medium of claim 61 wherein said program further includes connecting motions by (d) reading out plural time-series motion data stored said storage unit via said data bus, (b) connecting the plural time-series motion data to generate a series of time-series motion data, and (c) storing the series of time-series motion data in said storage unit via said data bus.