CONSTRUCTION SITE INSPECTION DOCUMENTATION METHOD AND SYSTEM
A method and system for construction inspection documentation with an inspection device moved at the construction along a trajectory, whilst capturing multiple inspection data sets. Thereby, there is automatically extracting first trajectory shape- and/or time-related attributes by evaluation of a shape and/or a timeline of the trajectory, automatically correlating extracted first attributes with corresponding second construction attributes provided by a stored construction plan or model, automatically aligning, and preferably also scaling, the trajectory using the construction plan or model as positional reference based on correlated first and second attributes. This enables automatically allocating a respective inspection data set to a position within the construction plan or model based on the aligned and scaled trajectory.
1 . A method for generating an inspection documentation of a construction, the method comprising:
moving an inspection device at the construction such that a trajectory develops, whilst capturing multiple inspection data sets of at least part of the construction along the trajectory,
automatically recording the trajectory of the moving based on motion and/or location data of the inspection device gathered on board/by a sensor of the inspection device whilst moving and
allocating a respective captured data set to a position of the trajectory, based on time stamps,
automatically extracting first trajectory shape- and/or time-related attributes by evaluation of a shape and/or a timeline of the trajectory,
automatically correlating extracted first attributes with corresponding second construction attributes provided by a stored construction plan or model,
automatically aligning, and scaling, the trajectory using the construction plan or model as positional reference based on correlated first and second attributes,
enabling automatically allocating a respective inspection data set to a position within the construction plan or model based on the aligned, and scaled, trajectory.
2 . The method according to claim 1 , further comprising automatically correcting the trajectory by adding, cutting and/or shifting trajectory points and/or segments based on constraints derived from the trajectory and/or construction model or plan.
3 . The method according to claim 2 , wherein constraints are linked to attributes.
4 . The method according to claim 1 , wherein first attributes are extracted according to pre-determined different classes of trajectory shape sections and/or trajectory motion sections.
5 . The method according to claim 1 , wherein second attributes comprise different classes of construction sections, therein at least one of:
classes of spaces,
classes of locations and/or dimensions of barriers, and/or
classes of openings.
6 . The method according to claim 1 , wherein attributes comprise:
patterns of shapes and/or motion patterns,
patterns of construction site sections, and/or
a graph representation.
7 . The method according to claim 4 , wherein patterns are derived from an association to a respective class.
8 . The method according to claim 3 , further comprising weighting of constraints depending on an attribute class.
9 . The method according to claim 1 , further comprising weighting a first attribute and/or a trajectory point and/or segment depending on an associated timestamp, in particular for attribute extraction, trajectory alignment and/or scaling and/or trajectory correction.
10 . The method according to claim 1 , wherein for the automatic aligning and/or scaling, there is consideration of:
the motion and/or location data of the inspection device and/or
a previous inspection documentation trajectory of the construction and/or at least one previous image of the construction from a previous inspection documentation.
11 . The method according to claim 1 , wherein the first attributes are extracted using a first machine learned algorithm, in particular comprising a first neural network, and in particular the second attributes are extracted from the construction plan or model using a second machine learned algorithm, in particular comprising a second neural network.
12 . The method according to claim 11 , wherein the first attributes are extracted using a first neural network.
13 . The method according to claim 12 , wherein the second attributes are extracted from the construction plan or model using a second machine learned algorithm.
14 . The method according to claim 13 , wherein the second attributes are extracted from the construction plan or model using a second neural network.
15 . The method according to claim 1 , further comprising extracting first attributes based on a 2D-planar presentation of the trajectory.
16 . The method according to claim 1 , wherein the inspection device comprises an opto-electronic 2D- and/or 3D-imaging device and the data sets comprise 2D- and/or 3D-images.
17 . The method according to claim 16 , wherein the motion and/or location data comprises at least part of the images.
18 . A construction inspection system for generating a construction inspection documentation, the system comprising:
a movable inspection device having:
motion and/or location detection means for recording a movement trajectory of the device and
inspection means for capturing inspection data sets of at least part of the construction along the trajectory and
an evaluation unit configured to allocate a respective captured data set to a position of the trajectory, based on time stamps,
wherein the evaluation unit is configured to:
extract first trajectory shape- and/or time-related attributes by evaluation of a shape and/or a timeline of the trajectory,
correlate extracted first construction site attributes with corresponding second construction site model attributes provided by a stored construction plan or model,
align, and scale, the trajectory using the construction plan or model as positional reference based on correlated first and second attributes,
allocate a respective inspection data set to a position within the construction plan or model based on the aligned, and scaled, trajectory.
19 . A computer program product comprising program code which is stored on a non-transitory machine-readable medium, and having computer-executable instructions for performing, when executed on a evaluation unit of a system, the method according to claim 1 .