Data processing method and apparatus
First Claim
1. A method for filtering a signal representative of features in a space of two dimensions, the signal comprising a plurality of data samples, each having a value representative of a predetermined parameter of the signal for a respective region of said space, the method comprising the steps of:
- (a) progressively filtering the signal at monotonically increasing filtering scales from S1 up to S2, the input signal to the first filtering scale being the original input signal to the filtering process, while the input signal to the or each subsequent filtering scale is the output signal from the previous filtering scale;
(b) at each filtering scale,i) identifying one or more connected-sets, each of either a plurality of adjoining data samples having substantially the same value or values lying in a predetermined range or a single respective data sample not in any other set, each set thus having a value equal to the value of the or all the data samples therein or, where the values lie in said range, the average of the values of the data samples therein; and
ii) selecting each of those collected-sets which has an area related in a predetermined way to the current filtering scale and which has a value that is a local maximum or a local minimum compared with those of adjacent connected sets,and for each selected connected-set,(c) identifying the adjacent connected-set whose value is closest to the current value of the selected set;
(d) so modifying the value of the selected set that the new value is equal to that of the adjacent connected-set identified in step (c) so that the adjacent connected-set becomes merged with the selected connected-set, thereby forming an extended connected-set of larger area; and
outputting an output signal from at least one of the filtering scales, wherein for at least one of such filtering scales, both connected-sets having local maxima and connected-sets having local minima are selected.
3 Assignments
0 Petitions
Accused Products
Abstract
In a data processing method and apparatus, an image signal representative of features in a space of two or more dimensions is progressively filtered at monotonically increasing filtering scales, the signal being input to the first filtering scale, and the input to the or each subsequent filtering scale is the output signal from the previous scale. Each input has a number of data samples each of which represents a respective region of the space, and at each filtering scale those samples are sorted into connected-sets, each of which contains a plurality of substantially equal value to data samples or a respective individual data sample not in other sets. The process then selects those connected-sets which contain data samples having values which are local maxima or local minima and which represent a composite region of a size related in a predetermined way to the current filtering scale. Each selected set is then effectively merged with the adjacent connected-set which contains data samples of a value closest to that of the sample(s) in the selected connected-set. This is achieved by changing the latter'"'"'s value to that of the samples in the adjacent connected set. The output signal for the filtering method is derived from the output of at least one of the filtering scales. In at least one of those scales, both connected-sets having local maxima and connected-sets having local minima are selected, thereby increasing the tolerance of the method to noise.
-
Citations
41 Claims
-
1. A method for filtering a signal representative of features in a space of two dimensions, the signal comprising a plurality of data samples, each having a value representative of a predetermined parameter of the signal for a respective region of said space, the method comprising the steps of:
-
(a) progressively filtering the signal at monotonically increasing filtering scales from S1 up to S2, the input signal to the first filtering scale being the original input signal to the filtering process, while the input signal to the or each subsequent filtering scale is the output signal from the previous filtering scale; (b) at each filtering scale, i) identifying one or more connected-sets, each of either a plurality of adjoining data samples having substantially the same value or values lying in a predetermined range or a single respective data sample not in any other set, each set thus having a value equal to the value of the or all the data samples therein or, where the values lie in said range, the average of the values of the data samples therein; and ii) selecting each of those collected-sets which has an area related in a predetermined way to the current filtering scale and which has a value that is a local maximum or a local minimum compared with those of adjacent connected sets, and for each selected connected-set, (c) identifying the adjacent connected-set whose value is closest to the current value of the selected set; (d) so modifying the value of the selected set that the new value is equal to that of the adjacent connected-set identified in step (c) so that the adjacent connected-set becomes merged with the selected connected-set, thereby forming an extended connected-set of larger area; and outputting an output signal from at least one of the filtering scales, wherein for at least one of such filtering scales, both connected-sets having local maxima and connected-sets having local minima are selected. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21)
-
-
22. A method for filtering a signal representative of features in a space of three dimensions, the signal comprising a plurality of data samples, each having a value representative of a predetermined parameter of the signal for a respective region of said space, the method comprising the steps of:
-
(a) progressively filtering the signal at monotonically increasing filtering scales from S1 up to S2, the input signal to the first filtering scale being the original input signal to the filtering process, while the input signal to the or each subsequent filtering scale is the output signal from the previous filtering scale; (b) at each filtering scale, i) identifying one or more connected-sets, each of either a plurality of adjoining data samples having substantially the same value or values lying in a predetermined range or a single respective data sample not in any other set, each set thus having a value equal to the value of the or all the data samples therein or, where the values lie in said range, the average of the values of the data samples therein; and ii) selecting each of those collected-sets which has a volume related in a predetermined way to the current filtering scale and which has a value that is a local maximum or a local minimum compared with those of adjacent connected sets, and for each selected connected-set, (c) identifying the adjacent connected-set whose value is closest to the current value of the selected set; (d) so modifying the value of the selected connected-set that the new value is equal to that of the adjacent connected-set identified in step (c) so that the adjacent connected-set becomes merged with the selected connected-set, thereby forming an extended connected-set of larger volume; and outputting an output signal from at least one of the filtering scales, wherein for at least one of such filtering scales, both connected-sets having local maxima and connected-sets having local minima are selected.
-
-
23. A method for filtering a signal representative of features in a space of two or more dimensions, the signal comprising a plurality of data samples each having a value representative of a predetermined parameter of the signal for a respective region of said space, the method comprising the following steps:
-
(a) progressively filtering the signal at monotonically increasing filtering scales from S1 up to S2, the input signal to the first filtering scale being the original input signal to the filtering process, while the input signal to the or each subsequent filtering scale is the output signal from the previous filtering scale; (b) at each filtering scale, i) identifying one or more connected-sets, each of either a plurality of adjoining data samples having substantially the same value or values lying in a predetermined range or a single respective data sample not in any other set, each set thus having a value equal to the value of the or all the data samples therein or, where the values lie in said range, the average of the values of the data samples therein; and ii) selecting each of those collected-sets which has a size and/or colour, a parameter or combination of parameters of which is related in a predetermined way to the current filtering scale and which has a value that is a local maximum or a local minimum compared with those of adjacent connected sets, and for each selected connected-set, (c) identifying the adjacent connected-set whose value is closest to the current value of the selected set; (d) so modifying the value of the selected connected-set that the new value is equal to that of the adjacent connected-set identified in step (c) so that the adjacent connected-set becomes merged with the selected connected-set, thereby forming an extended connected-set of larger size; and outputting the output signal from at least one of the filtering scales, wherein for at least one of such filtering scales, both connected-sets having local maxima and connected-sets having local minima are selected. - View Dependent Claims (24, 25, 26)
-
-
27. Apparatus for filtering a signal, the apparatus comprising:
-
(a) first circuit means for identifying connected-sets in an input signal; (b) comparator circuit means for determining whether the value of a connected-set is a local maximum or minimum by comparing that value with the value of adjacent connected-sets; (c) integration circuit means for determining one or more parameters related to the sizes of connected sets whose value is a local maximum or local minimum; (d) selection circuit means for comparing said parameter or parameters with a filtering scale thereby to determine whether such connected-sets require processing; (e) means for reading the values of the connected-sets that are adjacent to selected connected-sets; (f) further comparator means for determining which adjacent connected-set is closest in value to that of any selected connected-set; (g) means for modifying the value of a selected connected-set to that of the closest value adjacent connected-set; (h) output means for outputting a signal following filtering at a predetermined scale; and (i) means for determining when processing is finished. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41)
-
Specification