Artifact filtering using artificial intelligence
AssigneeFaro Technologies, Inc.
InventorsLouis Bergmann, Vadim Demkiv, Daniel Flohr
Life of this patent
Estimating the term…
Abstract
A system and a method for removing artifacts from a 3D coordinate data are provided. The system includes one or more processors and a measuring device. The one or more processors are operable to receive training data and train the 3D measuring device to identify artifacts by analyzing the training data. The one or more processors are further operable to identify artifacts in live data based on the training of the processor system. The one or more processors are further operable to generate clear scan data by filtering the artifacts from the live data and output the clear scan data.
Claims20
A method performed by a three-dimensional (3D) measuring device for improving scan data, the method comprising:
- storing in an electronic memory, for use by an artificial intelligence (AI) module having at least one processor, training data for identifying artifacts in scan data that correspond to moving objects using at least one of depth values and intensity values of individual pixels in the scan data;
- scanning an environment to collect live scan data;
- analyzing at least one of depth values and intensity values in a plurality of individual pixels in the live scan data to identify live artifacts using the training data;
- identifying an artifact in the live scan data based on the analysis of the training data;
- filtering the artifact from the live scan data to generate clear scan data;
- and
- outputting the clear scan data for generating a point cloud corresponding to the environment based on the clear scan data in place of the live scan data.
A method for generating and using trained scan data performed by at least one processor in communication with a three-dimensional (3D) measuring device, the method comprising:
- receiving training data to train a stored artificial intelligence (AI) module, the training data including a plurality of patterns corresponding to moving objects that have been identified by a human, the training data used to identify a pattern corresponding to a moving object according to changes in depth data, the pattern comprising an incomplete representation of the moving object during an acquisition of 3D coordinate data;
- capturing live scan data, including at least one of two-dimensional (2D) coordinate data and 3D coordinate data, using the 3D measuring device;
- identifying the pattern in live scan data, using the AI module, according to changes in the depth data in the live scan data generated by the 3D measuring device;
- generating clear scan data by filtering the pattern from the live scan data;
- and
- outputting the clear scan data in order to generate a point cloud corresponding to the environment based on the clear scan data without requiring a second separate scan of the environment.
A method for generating and using trained scan data performed by at least one processor in communication with a three-dimensional (3D) measuring device, the method comprising:
- receiving training data to train a stored artificial intelligence (AI) module, the training data including a plurality of patterns corresponding to moving objects that have been identified by a human, the training data used to identify a pattern corresponding to a moving object according to changes in intensity data, the pattern comprising an incomplete representation of the moving object during an acquisition of 3D coordinate data;
- capturing live scan data, including at least one of two-dimensional (2D) coordinate data and 3D coordinate data, using the 3D measuring device;
- identifying the pattern in live scan data, using the AI module, according to changes in the intensity data in the live scan data generated by the 3D measuring device;
- generating clear scan data by filtering the pattern from the live scan data;
- and
- outputting the clear scan data in order to generate a point cloud corresponding to the environment based on the clear scan data without requiring a second separate scan of the environment.
Dependencies are inferred from the claim text.