Data processing apparatus and method

• US 8,229,177 B2
• Filed: 05/31/2002
• Issued: 07/24/2012
• Est. Priority Date: 05/31/2001
• Status: Active Grant

• Current Assignee: Genkey Netherlands BV
• Original Assignee: Fountain Venture AS
• Inventors: Duffy, Dominic Gavan, Jones, Aled Wynne
• Primary Examiner(s): Allison, Andrae S
• Application Number:
  US10/478,827
• Publication Number:
  US 20040243356A1
• Time in Patent Office: 3,707 Days
• Field of Search: 382115-127, 902/3, 902/6, 380/44, 340/5.53, 340/5.83, 713/161
• US Class Current: 382/115
• CPC Class Codes: G06V 40/1365 Matching; Classification
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris
• CPC Class Codes: G06V 40/55 Performing matching on a personal external card, e.g. to avoid submitting reference information
• CPC Class Codes: H04L 2209/20 Manipulating the length of blocks of bits, e.g. padding or block truncation
• CPC Class Codes: H04L 2209/34 Encoding or coding, e.g. Huffman coding or error correction
• CPC Class Codes: H04L 2209/805 Lightweight hardware, e.g. radio-frequency identification [RFID] or sensor
• CPC Class Codes: H04L 9/3231 Biological data, e.g. fingerprint, voice or retina network architectures or network communication protocols for supporting authentication of entities using biometrical features in a packet data network H04L63/0861

Abstract

There is decribed an apparatus for generating a number from data originating from an analogue source, the apparatus comprising means for performing a set of the data and then processing, in accordance with stored processing instructions predetermined by a training process, the data to generate the number. During the training process, the sensitivity of the value of the generated number to variation in each of the measurement values is analyzed, and the process instructions are generated so that the processing of individual measurement values is modified to reduce this sensitivity. In this way, the repeatability of the generated number is improved.

Claims

1. A method of generating process information for a number generation process operable to generate a number representative of an analogue source from an analogue data set corresponding to the analogue source, the method comprising the steps of:

receiving for an analogue source at least one training analogue data set representative of said analogue source;

performing a set of measurements on each of the at least one training analogue data sets to obtain for each training analogue data set a corresponding set of measurement values suitable for processing by said number generation process;

for at least one measurement of the set of measurements;

analysing the stability of the number generated by the number generation process in response to change in the corresponding measurement value; and

determining at least one process instruction operable to modify the processing of the corresponding measurement value during the number generation process to improve the repeatability of numbers generated by the number generation process in case of change in said corresponding measurement value; and

generating processing information indicative of the at least one determined process instruction, whereby the repeatability of the number generated by the number generation process is improved for subsequent performances of the number generation process on analogue data sets corresponding to the analogue source by using said processing information.

2. The method according to claim 1, wherein the number generation process defines a sequence of numerical ranges, each numerical range having an upper numerical boundary and a lower numerical boundary, and generates the number representative of the analogue source in dependence upon the numerical range corresponding to a measurement value.

3. The method according to claim 2, wherein said analysing step comprises determining for at least one measurement of the set of measurements the position of the corresponding measurement value within said corresponding range, and determining a process instruction operable to vary the relative position between the corresponding measurement value and at least one of the upper and lower numerical boundaries to increase the distance between the corresponding measurement value and the closest range numerical boundary.

4. The method according to claim 3, wherein said analysing step comprises determining a process instruction comprising an offset value to be added to the measurement value.

5. The method according to claim 4, wherein the offset value is operable to position the corresponding measurement value in the middle of a numerical range.

6. The method according claim 2, wherein said receiving step comprises receiving a plurality of training analogue data sets,

wherein for at least one measurement of the set of measurements, the analysing step comprises statistically analysing the respective measurement values for the plurality of training analogue data sets, and determining a process instruction defining the sequence of numerical ranges associated with the respective measurement values in dependence upon said statistical analysis.

7. The method according to claim 6, wherein the analysing step is performed for plural measurements of the set of measurements, and a process instruction defining a respective different sequence of numerical ranges is determined for each of said analysed plurality of measurements.

8. The method according to claim 6, wherein said statistical analysis comprises calculating a value indicative of the standard deviation of the respective measurement values for the analysed measurement, wherein the determined process instruction sets the upper and lower numerical boundaries in dependence upon said value indicative of the standard deviation.

9. The method according to claim 1, wherein the number generation process comprises identifying a plurality of features within each training analogue data set, the plurality of features being identified by one or more of the set of measurements satisfying respective conditions.

10. The method according to claim 9, wherein said analysing step comprises analysing the stability of the identification of a feature to variation in the corresponding measurement values for the respective one or more of the set of measurements, and determining at least one process instruction operable to modify the identification of the feature to reduce the sensitivity of feature identifications to variation in said corresponding measurement values.

11. The method according to claim 10, wherein said receiving step comprises receiving a plurality of training analogue data sets, and wherein the analysing step is performed for a plurality of the set of measurements, and if a feature is not stably identified in all the plurality of the set of features, a process instruction excluding the feature.

12. The method according to claim 10, further comprising the step of indexing the identified features in dependence on respective measurement values associated with the features, wherein said analysing step comprises analysing the stability of the indexing to variation in the respective measurement values associated with the features, and determining at least one process instruction operable to modify the indexing of the features to improve repeatability of the indexing in case of variation in the respective measurement values.

13. The method according to claim 12, wherein the process instruction defines an indexing rule for a plurality of features.

14. The method according to claim 9, wherein said set of measurements comprises a first set of measurements for identifying features within the training analogue data sets, and a second set of measurements for measuring attributes of the identified features to obtain at least some of the measurement values.

15. The method according to claim 14, wherein the number generation process comprises classifying features in a plurality of categories in dependence upon one or more of the second set of measurements satisfying a condition.

16. The method according to claim 15, wherein said receiving step comprises receiving a plurality of training analogue data sets, and wherein the analysing step is performed for a plurality of the set of measurements, and if a feature is classified in different categories in different analogue data sets, a process instruction excluding the feature is determined.

17. The method according to claim 15, wherein said receiving step comprises receiving a plurality of training analogue data sets, and wherein the analysing step is performed for a plurality of the set of measurements, and if a feature is classified in three or more different categories in different analogue data sets, a process instruction excluding the feature is determined.

18. The method according to claim 17, wherein if a feature is classified in two different categories in different analogue data sets, a process instruction modifying the classification is determined.

19. The method according to claim 18, wherein the processing instruction comprises a classification rule for a feature.

20. The method according to claim 1, wherein at least one of the set of measurements comprises applying a transform function and the corresponding measurement value is given by the respective transform co-efficient.

21. The method according to claim 20, wherein the applied transform is a fast Fourier transform.

22. The method according to claim 20, wherein the applied transform is a wavelet transform.

23. The method according to claim 20, wherein the applied transform is a Gabor transform.

24. The method according to claim 1, wherein at least one of the set of measurements comprises a pattern recognition operation.

25. The method according to claim 1, wherein at least one of the set of measurements is performed by segregating the training analogue data set into plural parts, and individually measuring properties of the plural parts.

26. The method according to claim 25, wherein the average analogue data value is measured for each of the plural parts.

27. The method according to claim 1, further comprising applying the number generation process to generate the number representative of the analogue source.

28. The method according to claim 27, wherein the number generation process comprises the steps of:

deriving a plurality of data values and a plurality of error correction values using the set of measurement values;

combining the data values to generate the number representative of the analogue source; and

calculating error correction information using the error correction values and the data values.

29. The method according to claim 28, wherein the calculating error correction information step comprises processing each error correction value using an associated data value.

30. The method according to claim 29, wherein a bitwise exclusive-OR function is applied to each error correction value and the corresponding data value to generate the error correction information.

31. A method according to claim 27, wherein said number generation process comprises a process for generating a cryptographic key.

32. A method according to claim 27, wherein said number generation process comprises a process for generating a personal identification number.

33. The method according to claim 28, further comprising combining the error correction information and the processing information.

34. The method according to claim 1, further comprising performing a pre-processing step prior to performing said set of measurements, the pre-processing step comprising generating alignment data from the at least one received training analogue data set.

35. The method according to claim 34, wherein the receiving step comprises receiving a plurality of training analogue data sets, and wherein the pre-processing step comprises using the stored alignment data to align the information content of the plurality of training analogue data sets.

36. The method according to claim 34, wherein in the pre-processing step alignment data is generated by identifying a plurality of distinctive patterns within an analogue data set and storing the locations of the distinctive patterns.

37. The method according to claim 36, further comprising the storing the distinctive pattern in association with the corresponding location.

38. The method according to claim 34, further comprises normalising the training analogue data sets.

39. The method according to claim 1, wherein the analogue data source is a biological feature of an animal.

40. The method according to claim 39, wherein the analogue data source is the iris of an eye.

41. The method according to claim 39, wherein the analogue data source is a retina of an eye.

42. The method according to claim 39, wherein the analogue data source is a fingerprint.

43. The method according to claim 1, wherein the analogue data source is a photograph.

44. The method of generating a number representative of an analogue source, the method comprising the steps of:

generating process information in accordance with a method according to claim 1;

receiving an analogue data set representative of the analogue source;

performing the set of measurements on the analogue data set to obtain a set of measurement values; and

applying the number generation process using the set of measurement values corresponding to the analogue data set, in accordance with the processing instructions stored in said process information, to generate said number representative of the analogue source.

45. The method according to claim 1, wherein the at least one determined process instruction comprises an ambiguity resolution vector.

46. A method according to claim 1, wherein said number generation process comprises a process for generating a cryptographic key.

47. A method according to claim 1, wherein said number generation process comprises a process for generating a personal identification number.

48. A method according to claim 1, wherein said process instruction is operable, during the number generation process, to generate a modified measurement value from the measurement value.

49. A method of generating a number representative of an analogue source, the method comprising the steps of:

receiving an analogue data set representative of the analogue source;

performing a set of measurements on the analogue data set to obtain a set of measurement values;

retrieving process information from a data store; and

applying a number generation process using the set of measurement values, in accordance with the processing instructions stored in said process information, to generate said number representative of the analogue source;

wherein the process information is indicative of at least one determined process instruction for modifying the processing of said set of measurements to stabilize the number generated by the number generation process, thereby improving the repeatability of the number generated by the number generation process for the analogue source.

50. A method according to claim 49, wherein the number generation process defines a sequence of numerical ranges, each numerical range having an upper numerical boundary and a lower numerical boundary, and generates the number representative of the analogue source in dependence upon the numerical range corresponding to a measurement value.

51. A method according to claim 50, wherein said process information defines a process instruction operable to vary the relative position between a measurement value and at least one of the upper and lower numerical boundaries of the corresponding numerical range to increase the distance between the corresponding measurement value and the closest range numerical boundary.

52. A method according to claim 51, wherein said process instruction comprising an offset value which is added to the measurement value during the number generation process.

53. A method according to claim 52, wherein the offset value is operable to position the corresponding measurement value in the middle of a numerical range.

54. A method according to claim 50, wherein said process information comprises a process instruction defining the upper and lower boundaries for the sequence of numerical ranges associated with the respective measurement values.

55. A method according to claim 54, wherein the process information defines process instructions defining respective different sequence of numerical ranges at least two measurement values.

56. A method according to claim 54, wherein the determined process instruction sets the upper and lower numerical boundaries of the numerical ranges for a corresponding measurement value in dependence upon a predetermined standard deviation for the corresponding measurement value.

57. A method according to claim 49, wherein the number generation process comprises identifying a plurality of features within the analogue data set, the plurality of features being identified by one or more of the set of measurements satisfying respective conditions.

58. A method according to claim 57, wherein said process information comprises at least one process instruction operable to modify the identification of a feature.

59. A method according to claim 58, wherein said process information comprises a process instruction excluding a feature associated with one or more measurement values.

60. A method according to claim 57, wherein said set of measurements comprises a first set of measurements to obtain a first set of measurement values for identifying features within the analogue data sets, and a second set of measurements for measuring attributes of the identified features to obtain a second set of the measurement values.

61. A method according to claim 60, wherein the number generation process comprises classifying features in a plurality of categories in dependence upon one or more of the second set of measurements satisfying respective conditions.

62. A method according to claim 61, wherein the process information defines a process instruction excluding an identified feature.

63. A method according to claim 61, wherein the process information defines a process instruction modifying the classification of an identified feature.

64. A method according to claim 63, wherein the processing instruction comprises a classification rule for a feature.

65. A method according to claim 58, wherein the number generation process comprises the step of indexing the identified features in dependence on respective measurement values associated with the features, wherein the process information comprises at least one process instruction operable to modify the indexing of the features.

66. A method according to claim 65, wherein the process instruction defines an indexing rule for a plurality of features.

67. A method according to claim 49, wherein at least one of the set of measurements comprises applying a transform function and the corresponding measurement value is given by the respective transform co-efficient.

68. A method according to claim 67, wherein the applied transform is a fast Fourier transform.

69. A method according to claim 67, wherein the applied transform is a wavelet transform.

70. A method according to claim 67, wherein the applied transform is a Gabor transform.

71. A method according to claim 49, wherein at least one of the set of measurements comprises a pattern recognition operation.

72. A method according to claim 49, wherein at least one of the set of measurements is performed by segregating the analogue data set into plural parts, and individually measuring properties of the plural parts.

73. A method according to claim 72, wherein an average analogue data value is measured for each of the plural parts.

74. A method according to claim 49, wherein the number generation process comprises the steps of:

deriving a plurality of data values and a plurality of error correction values using the set of measurement values, each data value being associated with at least one error correction value;

processing the each data value and the corresponding at least one error correction value in accordance with stored error correction information to generate a plurality of adjusted data values; and

combining the adjusted data values to generate the number representative of the analogue source.

75. A method according to claim 74, wherein a bitwise exclusive-OR function is applied to each error correction value and a respective binary number stored in the error correction information.

76. A method according to claim 49, further comprising performing a pre-processing step prior to performing said set of measurements, the pre-processing step comprising generating aligning the received analogue data set using stored alignment data.

77. A method according to claim 76, wherein the alignment data identifies reference positions corresponding to a plurality of distinctive patterns within the analogue data set.

78. A method according to claim 77, wherein the alignment data further comprises the plurality of distinctive patterns stored in association with the corresponding locations.

79. A method according to claim 76, wherein the pre-processing step further comprises normalising the received analogue data set.

80. A method according to claim 49, wherein the analogue data source is a biological feature of an animal.

81. A method according to claim 80, wherein the analogue data source is the iris of an eye.

82. A method according to claim 80, wherein the analogue data source is a retina of an eye.

83. A method according to claim 80, wherein the analogue data source is a fingerprint.

84. A method according to claim 49, wherein the analogue data source is a photograph.

85. A storage device storing instructions including instructions for causing a programmable apparatus to perform a method according to claim 49.

86. A signal conveying instructions including instructions for causing a programmable apparatus to perform a method according to claim 49.

106. The apparatus according to claim 86, wherein said measurer is operable to apply a transform function to a received training analogue data set to determine a corresponding measurement value.

107. The apparatus according to claim 106, wherein the measurer is operable to apply a fast Fourier transform.

108. The apparatus according to claim 106, wherein the measurer is operable to apply a wavelet transform.

109. The apparatus according to claim 106, wherein the measurer is operable to apply a Gabor transform.

87. An apparatus for generating process information for a number generation process operable to generate a number representative of an analogue source from an analogue data set corresponding to the analogue source, the apparatus comprising:

a receiver operable to receive for an analogue source at least one training analogue data set representative of said analogue source;

a measurer operable to perform a set of measurements on each of the at least one training analogue data sets to obtain for each training analogue data set a corresponding set of measurement values suitable for processing by said number generation process;

an analyser operable to;

analyse, for at least one measurement of the set of measurements, the stability of the number generated by the number generation process in response to change in the corresponding measurement value; and

to determine at least one process instruction operable to modify the processing of the corresponding measurement value during the number generation process to improve the repeatability of numbers generated by the number generation process in case of change in said corresponding measurement value; and

an information generator operable to generate processing information indicative of the at least one determined process instruction, whereby the repeatability of the number generated by the number generation process is improved for subsequent performances of the number generation process on analogue data sets corresponding to the analogue source by using the processing information.

88. The apparatus according to claim 87, wherein the number generation process defines a sequence of numerical ranges, each numerical range having an upper numerical boundary and a lower numerical boundary, and generates the number representative of the analogue source in dependence upon the numerical range corresponding to a measurement value.

89. The apparatus according to claim 88, wherein said analyser comprises a determiner operable to determine for at least one of the set of measurements the position of the corresponding measurement value within said corresponding range, and operable to determine a process instruction operable to vary the relative position between the corresponding measurement value and at least one of the upper and lower numerical boundaries to increase the distance between the corresponding measurement value and the closest range numerical boundary.

90. The apparatus according to claim 89, wherein said analyser is arranged to determine a process instruction comprising an offset value to be added to the measurement value.

91. The apparatus according to claim 90, wherein the offset value is operable to position the corresponding measurement value in the middle of a numerical range.

92. The apparatus according to claim 88, wherein said analyser is operable to analyse a plurality of sets of measurement values corresponding respectively to a plurality of received training analogue data sets and for at least one of the set of measurements, to analyse statistically the respective measurement values for the plurality of training analogue data sets, and to determine a process instruction defining the sequence of numerical ranges associated with the respective measurement values in dependence upon said statistical analysis.

93. The apparatus according to claim 92, wherein the analyser is operable to analyse a plurality of measurements within the same set of measurements, and to determine process instructions defining respective different sequences of numerical ranges for each of said analysed plurality of measurements.

94. The apparatus according to claim 92, wherein said analyser is operable to perform said statistical analysis by calculating a value indicative of the standard deviation of the respective measurement values for the analysed measurement, wherein the analyser is operable to determine the upper and lower numerical boundaries of a numerical range in dependence upon said value indicative of the standard deviation.

95. The apparatus according to claim 87, wherein the number generation comprises identifying a plurality of features within each training analogue data set, the plurality of features being identified by one or more measurements of a set of measurements satisfying respective conditions.

96. The apparatus according to claim 95, wherein said analyser is operable to analyse the stability of the identification of a feature in case of variation in the corresponding measurement values, and to determine at least one process instruction operable to modify the identification of the feature to improve the repeatability of feature identifications to variation in said corresponding measurement values.

97. The apparatus according to claim 96, wherein the analyser is operable to analyse a plurality of sets of measurements corresponding to a plurality of training analogue data sets, and to determine a process instruction excluding a feature if the excluded feature is not identified in all the plurality of the set of measurements.

98. The apparatus according to claim 96, further comprising an indexer operable to index the identified features in dependence on respective measurement values associated with the features, wherein said analyser is operable to analyse the stability of the indexing to variation in the respective measurement values associated with the features, and to determine at least one process instruction operable to modify the indexing of the features to improve the repeatability of the indexing to variation in the respective measurement values.

99. The apparatus according to claim 98, wherein the process instruction defines an indexing rule for a plurality of features.

100. The apparatus according to claim 96, wherein said measurer is operable to measure a first set of measurements identifying features within the or each training analogue data set, and a second set of measurements measuring attributes of the identified features to obtain.

101. The apparatus according to claim 100, wherein the number generation comprises classifying features in a plurality of categories in dependence upon one or more of the second set of measurements satisfying a condition.

102. The apparatus according to claim 101, wherein the analyser is operable to analyse a plurality of sets of measurements corresponding to a respective plurality of training analogue data sets, and to determine a process instruction excluding a feature if the excluded feature is classified in different categories for different analogue data sets.

103. The apparatus according to claim 101, wherein the analyser is operable to perform a plurality of sets of measurements corresponding to a respective plurality of training analogue data sets, and to determine a process instruction excluding a feature if the excluded feature is classified in three or more different categories in different analogue data sets.

104. The apparatus according to claim 102, wherein if a feature is classified in two different categories in different analogue data sets, the analyser is operable to determine a process instruction modifying the classification.

105. The apparatus according to claim 104, wherein the analyser is operable to define a processing instruction comprises a classification rule for a feature.

110. The apparatus according to claim 87, wherein the measurer is operable to perform a pattern recognition operation.

111. The apparatus according to claim 87, wherein the measurer is operable to segregate the or each training analogue data set into plural parts, and to measure individual properties of the plural parts.

112. The apparatus according to claim 111, wherein the measurer is operable to measure an average analogue data value for each of the plural parts.

113. The apparatus according to claim 87, further comprising a number generator operable to generate said number representative of the analogue source.

114. The apparatus according to claim 113, wherein the number generator comprises:

a deriver operable to derive a plurality of data values and a plurality of error correction values using the set of measurement values;

a combiner operable to combine the data values to generate the number representative of the analogue source; and

a calculator operable to calculate error correction information using the error correction values and the data values.

115. The apparatus according to claim 114, wherein the calculator comprises a processor operable to process each error correction value using an associated data value.

116. The apparatus according to claim 115, wherein said processor is operable to apply a bitwise exclusive-OR function to each error correction value and the corresponding data value to generate the error correction information.

117. The apparatus according to claim 114, further comprising a combiner operable to combine the error correction information and the processing information.

118. The apparatus according to claim 87, further comprising a pre-processor operable to perform a pre-processing operation to generate alignment data from the at least one received training analogue data set.

119. The apparatus according to claim 118, wherein the pre-processor comprises an aligner operable to align the information content of the plurality of training analogue data sets using the stored alignment data.

120. The apparatus according to claim 118, wherein the pre-processor is operable to generate alignment data by identifying a plurality of distinctive patterns within an analogue data set and storing the locations of the distinctive patterns.

121. The apparatus according to claim 120, further comprising a data store operable to store the distinctive pattern in association with the corresponding location.

122. The apparatus according to claim 118, further comprises a normaliser operable to normalise the training analogue data sets.

123. The apparatus according to claim 87, further comprising an imager operable to image a biological feature of an animal to generate the or each training analogue data set.

124. The apparatus according to claim 123, wherein the imager is adapted to image the iris of an eye.

125. The apparatus according to claim 123, wherein the imager is adapted to image a retina of an eye.

126. The apparatus according to claim 123, wherein the imager is a fingerprint sensor.

127. The apparatus according to claim 87, wherein the imager is adapted to image a photograph.

128. An apparatus for generating a number representative of an analogue source, the apparatus comprising:

a receiver operable to receive an analogue data set representative of the analogue source;

a measurer operable to perform a set of measurements on the analogue data set to obtain a set of measurement values;

a data store for storing process information indicative of processing instructions operable to modify selectively the processing of the set of measurement values to improve the stability of the generated number in case of variations in the set of measurement values obtained from different analogue data sets representative of the analogue source; and

a number generator operable to generate said number representative of the analogue source using the set of measurement values and using said processing instructions to improve the repeatability of the number generated by the number generation process for the analogue source.

129. An apparatus according to claim 128, wherein the number generator is operable to define a sequence of numerical ranges, each numerical range having an upper numerical boundary and a lower numerical boundary, and to generate the number representative of the analogue source in dependence upon the numerical range corresponding to a measurement value.

130. An apparatus according to claim 129, wherein said number generator is operable to vary the relative position between a measurement value and at least one of the upper and lower numerical boundaries of the corresponding numerical range in accordance with a processing instruction indicated by the process information.

131. An apparatus according to claim 130, wherein the number generator is operable to add an offset value to the measurement value.

132. An apparatus according to claim 131, wherein the offset value is operable to position the corresponding measurement value in the middle of a numerical range.

133. An apparatus according to 129, wherein said number generator is operable to define the upper and lower boundaries for the sequence of numerical ranges associated with the respective measurement values in accordance with process instructions indicated by the process information.

134. An apparatus according to claim 133, wherein the number generator is operable to define respective different sequences of numerical ranges for at least two measurement values in accordance with process instructions indicated by the process information.

135. An apparatus according to claim 133, wherein the determined process instruction sets the upper and lower numerical boundaries of the numerical ranges for a corresponding measurement value in dependence upon a predetermined standard deviation for the corresponding measurement value.

136. An apparatus according to claim 128, wherein the number generator comprises an identifier operable to identify a plurality of features within the analogue data set, the plurality of features being identified by one or more of the set of measurements satisfying respective conditions.

137. An apparatus according to claim 136, wherein said number generator is operable to modify the identification of a feature in accordance with a process instruction indicated by the process information.

138. An apparatus according to claim 137, wherein said number generator is operable to exclude a feature in accordance with a process instruction indicated by the process information.

139. An apparatus according to claim 137, wherein the number generator comprises an indexer operable to index the identified features in dependence on respective measurement values associated with the features, wherein the number generator is operable to modify the indexing of the features in accordance with a process instruction indicated by the process information.

140. An apparatus according to claim 139, wherein the process instruction defines an indexing rule for a plurality of features.

141. An apparatus according to claim 128, wherein said measurer is operable to perform a first set of measurements to obtain a first set of measurement values for identifying features within the analogue data sets, and a second set of measurements for measuring attributes of the identified features to obtain a second set of the measurement values.

142. An apparatus according to claim 141, wherein the number generator comprises a classifier operable to classify features in a plurality of categories in dependence upon one or more of the second set of measurements satisfying respective conditions.

143. An apparatus according to claim 142, wherein the number generator is operable to exclude an identified feature in accordance with a process instruction indicated by the process information.

144. An apparatus according to claim 142, wherein the number generator is operable to modify a classification of an identified feature in accordance with a process instruction indicated by the process information.

145. An apparatus according to claim 144, wherein the processing instruction comprises a classification rule for a feature.

146. An apparatus according to claim 128, wherein said measurer is operable to apply a transform function to a received training analogue data set to determine a corresponding measurement value.

147. An apparatus according to claim 146, wherein the measurer is operable to apply a fast Fourier transform.

148. An apparatus according to claim 146, wherein the measurer is operable to apply a wavelet transform.

149. An apparatus according to claim 146, wherein the measurer is operable to apply a Gabor transform.

150. An apparatus according to claim 149, further comprising a pre-processor operable to perform a pre-processing operation to generate alignment data from the at least one received training analogue data set.

151. An apparatus according to claim 150, wherein the pre-processor comprises an aligner operable to align the information content of the plurality of training analogue data sets using the stored alignment data.

152. An apparatus according to claim 150, wherein the pre-processor is operable to generate alignment data by identifying a plurality of distinctive patterns within an analogue data set and storing the locations of the distinctive patterns.

153. An apparatus according to claim 152, further comprising a data store operable to store the distinctive pattern in association with the corresponding location.

154. An apparatus according to claim 128, wherein the measurer is operable to perform a pattern recognition operation.

155. An apparatus according to claim 128, wherein the measurer is operable to segregate the or each training analogue data set into plural parts, and to measure individual properties of the plural parts.

156. An apparatus according to claim 155, wherein the measurer is operable to measure an average analogue data value for each of the plural parts.

157. An apparatus according to claim 128, wherein the number generator comprises:

a deriver operable to derive a plurality of data values and a plurality of error correction values using the set of measurement values;

a combiner operable to combine the data values to generate the number representative of the analogue source; and

a calculator operable to calculate error correction information using the error correction values and the data values.

158. An apparatus according to claim 157, wherein the calculator comprises a processor operable to process each error correction value using an associated data value.

159. An apparatus according to claim 158, wherein said processor is operable to apply a bitwise exclusive-OR function to each error correction value and the corresponding data value to generate the error correction information.

160. An apparatus according to claim 157, further comprising a combiner operable to combine the error correction information and the processing information.

161. An apparatus according to claim 160, further comprising a normaliser operable to normalise the training analogue data sets.

162. An apparatus according to claim 128, further comprising an image operable to image a biological feature of an animal to generate the or each training analogue data set.

163. An apparatus according to claim 162, wherein the imager is adapted to image the iris of an eye.

164. An apparatus according to claim 162, wherein the imager is adapted to image a retina of an eye.

165. An apparatus according to claim 162, wherein the imager is a fingerprint sensor.

166. An apparatus according to claim 128, wherein the imager is adapted to image a photograph.