Data Processing (Measurement System)

The data obtained from the 3D scanner is processed for input to the trained model during deployment.

class dlmfg.core.wls400a_system.GetInferenceData[source]

Inference Data Class

Import required files to deploy model on measurement systems

data_pre_processing(measurement_data, voxel_channels=1)[source]

Process measurement data and impute missing values

Parameters
  • measurement_data (str (required)) – file name of the tab delimited file given as output from CoreviewAM

  • voxel_channels (int (required)) – The number of voxel channels that can be extracted from the the measurement file

Returns

numpy array of the node deviations (this is similar to what is obtained from the VRM software )

Return type

numpy.array [1*nodes]

load_mapping_index(index_file)[source]

Import mapping index used to map nodes to voxel locations from the file structure

Parameters

index_file – Path to the index file and the index file name

Returns

numpy array of voxel mapping index for each node

Return type

numpy.array [point_dim,3]

load_measurement_file(measurement_file_name)[source]

Import measurement file on which the model is to be deployed

Parameters

measurement_file_name (str (required)) – file name of the tab delimited file given as output from CoreviewAM

Returns

numpy array of the file after eliminating meta data information

Return type

numpy.array

voxel_mapping(y_dev_data_filtered, voxel_point_index, point_dim, voxel_dim, voxel_channels)[source]

Map the node deviations to voxel structure for input to the 3D CNN model

Parameters
  • y_dev_data_filtered (numpy.array (required)) – numpy array of the node deviations

  • voxel_point_index (numpy.array [nodes*3] (required)) – mapping index

  • point_dim (int (required)) – the number of nodes

  • point_dim – the number of nodes

  • voxel_dim (int (required)) – The resolution of the voxel

Returns

voxel_dev_data (input to the 3D CNN model)

Return type

np_array [1*voxel_dim,voxel_dim,voxel_dim,voxel_channels]