초록 ▼

The invention relates to a method and system for processing image data obtained from scanning a network infrastructure for the detection and analysis of specific objects of interest, such as powerlines and other overhead conductors, or similar structures. The image data comprises a plurality of co-o

The invention relates to a method and system for processing image data obtained from scanning a network infrastructure for the detection and analysis of specific objects of interest, such as powerlines and other overhead conductors, or similar structures. The image data comprises a plurality of co-ordinate points in three-dimensional space, and in order to identify conductors in the network infrastructure, the method involves analyzing the co-ordinate points to identify sets of co-ordinate points indicative of a set of substantially parallel lines, and analyzing the co ordinate points on the basis of said identified substantially parallel lines thereby to allow identification and mapping of hanging catenaries representing said conductors.

대표청구항 ▼

1. A method of processing image data obtained from scanning a network infrastructure in order to identify conductors in the network, said image data comprising a plurality of co-ordinate points in three-dimensional space, the method comprising the steps of: (a) dividing the image data into a grid of

1. A method of processing image data obtained from scanning a network infrastructure in order to identify conductors in the network, said image data comprising a plurality of co-ordinate points in three-dimensional space, the method comprising the steps of: (a) dividing the image data into a grid of cells, each representing a volume within the three-dimensional space and, for each cell: (i) identifying sets of co-ordinate points from potential co-ordinate points within the cell, each said set being indicative of a plurality of identified substantially parallel vertical planes, each said plane representing a possible hanging catenary;(ii) identifying a plurality of individual subsets of coordinate points from each set, each subset representing an identified vertical plane in the plurality of substantially parallel vertical planes;(iii) analysing the distribution of co-ordinate points of a vertical plane to identify one or more vertical plane segments, including allocating to each vertical plane segment the coordinate points used in identifying that vertical plane segment;(b) for the one or more vertical plane segments generated by the analysis of all cells, aggregating the vertical plane segments into one or more linear runs, on the basis of one or more prescribed aggregation criteria; and(c) for the one or more linear runs generated: (i) defining co-ordinate points within each linear run as belonging to a particular span and circuit; and(ii) defining co-ordinate points within each circuit as belonging to a particular conductor. 2. A method according to claim 1, wherein the step of identifying the sets of co-ordinate points from the potential co-ordinate points within the cell comprises applying a Hough transformation to the potential co-ordinate points in a projection of those points onto a horizontal plane. 3. A method according to claim 2, wherein the step of applying a Hough transformation includes analysing potential co-ordinate points in an R,θ accumulator space, and incrementing the accumulator for each θ and R for each co-ordinate point that lies within that vertical plane, such that on completion of analysis of all potential co-ordinate points, the accumulators represent the number of coordinate points within each vertical plane, θ represents the angular orientation of the plane in geographical terms and R represents the distance to the plane from a prescribed position in the cell. 4. A method according to claim 3, wherein the plurality of substantially parallel vertical planes are identified by summing the square of the accumulator within a tolerance band of each R,θ pair. 5. A method according to claim 2, further including the step of identifying sets of accumulators having peak values of similar θ with prescribed R separation, each said set of accumulators representing substantially parallel vertical planes separated by a prescribed distance. 6. A method according to claim 2, wherein the step of identifying a plurality of individual subsets of coordinate points comprises identifying individual local maxima in Hough accumulator space around an R,θ representing an identified central vertical plane. 7. A method according to claim 1, wherein the step of analysing the distribution of coordinate points of a vertical plane comprises: (a) applying statistical analysis to the distribution of co-ordinate points within each identified vertical plane, so as to provide a set of ranked vertical plane segments; and(b) allocating the co-ordinate points from the highest ranked plane segment and removing the allocated co-ordinate points from the set of potential co-ordinate points. 8. A method according to claim 7, wherein the step of applying statistical analysis to the distribution of co-ordinate points within each identified vertical plane is iteratively repeated on a reduced set of co-ordinate points with respect to the remaining vertical plane segments, with progressively lower rankings being reassigned to the remaining planes, to identify the successively next prominent vertical plane segment. 9. A method according to claim 8, wherein the step of analysing the distribution of co-ordinate points within each identified vertical plane is repeated until no vertical plane segment remains whose ranking meets a prescribed threshold. 10. A method according to claim 7, wherein the vertical plane segment ranking relates to parameters including: linearity of co-ordinate point sets; uniformity of distribution of co-ordinate point sets; gaps within co-ordinate point sets. 11. A method according to claim 1, wherein the step of aggregating the vertical plane segments into one or more linear runs comprises analysing the alignment and separation of the vertical plane segments and, as a result of the analysis, grouping plane segments into a set considered as a run. 12. A method according to claim 11, wherein the step of analysing the alignment and separation of vertical plane segments comprises: (a) applying a ranking to the vertical plane segments;(b) allocating the highest ranked vertical plane segment to a first run with a nominal zero width; and(c) adding to the first run all vertical plane segments which meet one or more prescribed run inclusion criteria. 13. A method according to claim 12, wherein the one or more run inclusion criteria includes: whether the difference in orientation between the vertical plane segment and the first run is within a prescribed range; whether the vertical plane segment is within a prescribed distance from the first run. 14. A method according to claim 12, wherein the step of adding vertical plane segments to said first run includes adjusting the parameters of the first run in accordance with the new vertical plane segment. 15. A method according to claim 12, wherein the step of adding vertical plane segments to said first run is repeated until no further vertical plane segments can be added to the first run. 16. A method according to claim 12, wherein the step of analysing the alignment and separation of the vertical plane segments is repeated to establish second and further runs. 17. A method according to claim 1, wherein the step of defining co-ordinate points within each linear run as belonging to a particular span and circuit comprises: (a) projecting all co-ordinate points allocated to vertical planes within a particular run to a centreline plane of the run; and(b) analysing the projected points in the 2D space defined by that vertical plane. 18. A method according to claim 17, wherein the step of analysing the projected points is conducted by way of a Hough analysis in 3D Hough space, so as to identify catenaries and sets of vertically separated catenaries. 19. A method according to claim 17, wherein the Hough analysis involves identifying catenaries and grouping the identified catenaries into catenary sets based on the location of the catenary vertex, said catenary vertex being defined in terms of D0 (distance of vertex along the run) and H (height of vertex). 20. A method according to claim 19, wherein catenaries with proximate D0 are allocated to the same span, whilst catenaries allocated to the same span but with H differing by at least a prescribed value are allocated to different circuits. 21. A method according to claim 1, wherein the step of defining co-ordinate points within each circuit as belonging to a particular conductor comprises establishing a set of vertical planes each containing a single catenary, and thus a single conductor. 22. A method according to claim 21, wherein the step of defining co-ordinate points within each circuit as belonging to a particular conductor is repeated on all points allocated to each individual conductor vertical plane, to result in defined single catenaries representing individual conductors. 23. A method according to claim 1, including the step of uniquely identifying and characterising each individual conductor. 24. A method according to claim 23, wherein the step of uniquely identifying and characterising each individual conductor comprises applying a least squares regression technique to determine the best fit catenary parameters. 25. A method according to claim 1, wherein the positions of end points of the particular conductors is determined by calculating intersect points between immediately adjacent conductors. 26. A method according to claim 25, further including the step of identifying powerline poles by analysing co-ordinate points not associated with conductors in the proximity of the end points of the conductors. 27. A method according to claim 26, wherein a priori and/or a posteriori information regarding the powerline poles is used to refine the positions of end points of the conductors. 28. A method of processing image data obtained from scanning a network infrastructure in order to identify conductors, said image data comprising a plurality of co-ordinate points in three-dimensional space, the method including the steps of: (a) analysing the co-ordinate points to identify sets of co-ordinate points indicative of a plurality of substantially parallel lines; and(b) analysing each of the sets identified in said step (a) to identify co-ordinate points indicative of a line from said plurality of substantially parallel lines to thereby identify and map hanging catenaries representing said conductors. 29. A method according to claim 28, wherein the substantially parallel lines represent either substantially parallel vertical planes or substantially parallel vertically separated hanging catenaries. 30. A non-transitory computer-readable medium storing a computer program including logic functionality which, when executed by a processor, implements, when provided with a scanned image data set, the method according to claim 1, and outputs data representative of identified conductors in a network infrastructure. 31. A computer-based system for processing scanned image data to identify conductors in a network infrastructure, the system including: (a) data input means to receive scanned image data;(b) a processing unit configured to carry out the method according to claim 1; and(c) output means to provide data representative of identified conductors in a network infrastructure. 32. A non-transitory computer-readable medium storing a computer program including logic functionality which, when executed by a processor, implements, when provided with a scanned image data set, the method according to claim 28, and outputs data representative of identified conductors in a network infrastructure. 33. A computer-based system for processing scanned image data to identify conductors in a network infrastructure, the system including: (a) data input means to receive scanned image data;(b) a processing unit configured to carry out the method according to claim 28; and(c) output means to provide data representative of identified conductors in a network infrastructure.