308#ifdef NEW_VERSION_SDINET1
312 SB_NETWORK_TYPE_SDINET1 = SB_NETWORK_TYPE_SDINET1
339 SB_NETWORK_FREEZE_MODE_END = SB_NETWORK_FREEZE_MODE_END
Blob analysis parameters Class that wraps the structure sb_t_blob_par
Defines computing device types that wraps the sb_t_devices_par structure
array< int > id
Identifier of the devices to be used
SbDeviceType type
Device computational type
SbFrameworkType framework
Identifier of the framework to be used
Parameters Class that wraps the sb_t_par structure. You must call the Dispose() method to free all th...
array< unsigned char, 2 > collaborations
Matrix of the models collaborations
static String FormatImageCircularityType(SbImageCircularityType value)
Formats the image circularity type
SbProjectType project_type
Project type
SbError RemoveLevel(String^ model_name, float scale)
Removes the l-th level from the structure of the project parameters.
SbBlobPar blob_par
Blob analysis parameters. Used only for Surface projects
static String FormatImageBordersExtensionMode(SbImageBordersExtensionMode value)
Formats the Deep Learning image borders extension mode
static String FormatProjectType(SbProjectType value)
Formats the project type
static String FormatFloatingPointPrecisionType(SbFloatingPointPrecisionType value)
Formats the floating point precision
SbFloatingPointPrecisionType floating_point_precision
Floating point precision used to run operations during detection.
int GetCollaboration(String^ model_name1, String^ model_name2)
Gets collaboration between two models.
SbError RemoveModel(String^ model_name)
Removes the model from the parameter structure.
String Format()
Formats the parameters string.
static String FormatNetworkType(SbNetworkType value)
Formats the Deep Learning Network type
SbError SetCollaboration(String^ model_name1, String^ model_name2, bool value)
Sets collaboration between two models.
array< SbParModel^> models
Array of the models parameters
SbInterpolationMode resize_mode
Interpolation mode used to resize the images.
SbError AddModel(String^ model_name)
Adds the model to the parameter structure.
static String FormatLossFnType(SbLossFnType value)
Formats the Deep Learning Loss function type
static String FormatTilingMode(SbTilingMode value)
Formats the Deep Learning tiling mode
static String FormatPerturbationType(SbPerturbationType value)
Formats the Deep Learning SVL image perturbation type
SbParSl sl
Shallow Learning modules parameters
static String FormatSvlParOptimizationMode(SbSvlParOptimizationMode value)
Formats the Shallow Learning SVL optimization mode
SbSvlPar svl
SVL parameters
SbDevicesPar devices
Devices used for inference.
static String FormatPerturbationMode(SbPerturbationMode value)
Formats the Deep Learning SVL image perturbation mode
int surface_blob_analysis
Enable the surface blob analysis
SbError AddLevel(String^ model_name, float scale)
Adds the level to the parameter structure.
SbRgba color_opt
Optional all models color. Used for painting
int num_threads
Num threads to be used for the detection
Level parameters Class that wraps the sb_t_level_par structure
bool enabled
Enabling flag of the level.
Model parameters class that wraps the sb_t_par_model structure
int num_occurrences
Num occurrences.
SbSize obj_min_distance
Objects min distance
SbRgba color_opt
Paint color of the model optional defect. Used only for Surface projects
SbSize obj_stride_fine
Fine search step.
array< SbPerturbation^> perturbations
Array of the perturbations of the model
float defect_area_percentage
String description
Model description
SbSize obj_stride_coarse
Coarse search step.
float defect_area_threshold
SbSize obj_size
Object size
bool enabled
Model is enabled
SbRgba color
Paint color of the model defect. Used only for Surface projects
array< SbParLevel^> levels
Array of the detection levels associated to the model
Shallow Learning modules parameters class that wraps the sb_t_par_sl structure
float speed_boost
Detection speed boost.
int detection_out_of_roi
Detection near or partially outside the analysis roi.
Perturbations Class that wraps the sb_t_perturbation structure
SbRange angle_range
Angular range, in degrees, for random rotation.
int num_synthetic_samples
Number of synthetic sample to generated with a random angle in the range sb_t_par_perturbation::angle...
int flip_horizontal
Flip around y-axis. 2th operation.
int flip_vertical
Flip around x-axis. 3th operation.
int angle
Rotation angle, in degree, of the sample. 1th operation.
Range value class that wraps the sb_t_range_flt structure
Range value class that wraps the sb_t_range structure
rgba class that wraps the sb_t_rgba structure
Size class that wraps the sb_t_size structure
SVL parameters that configure the Shallow Learning training, it wraps the sb_t_svl_sl_par structure
SbSvlDlTilingPar tiling_par_height
Height tiling configuration.
float learning_rate
Learning rate.
String pre_training_file
Network weights file path with extension SB_PRE_TRAINING_EXT.
float validation_percentage
Validation percentage.
SbSvlDlParNetwork network
Network parameters.
SbSvlDlParPerturbation perturbations
Perturbations for deep learning training.
int save_best
At the end of the training, the best internal parameters configuration is recovered.
SbImageBordersExtensionMode borders_extension_mode
Loss Image borders extension mode.
int batch_size
Size of the batch used during SVL.
SbSvlDlTilingPar tiling_par_width
Width tiling configuration.
SbLossFnType loss_fn
Loss function.
int num_epochs
Number of epochs.
Deep Learning network parameters class that wraps the sb_t_svl_dl_par_network structure
SbNetworkType type
Network type.
SbSize input_size
Network input size.
int n_channels
Network input channels.
float features_multiplier
Features multiplier factor of the network.
SbImageFormat image_format
Network input image format.
SbNetworkFreezeMode freeze_mode
Freeze mode to apply on network parameters during SVL.
Describes the perturbation of the image / defect, it wraps the sb_t_svl_dl_par_perturbation structure
SbPerturbationMode mode
Select the perturbation mode.
SbPerturbationType type
Select the perturbation type.
int flip_horizontal
Flip around y-axis.
float delta_brightness
Maximum delta of brightness to apply to the image.
int flip_vertical
Flip around x-axis.
SbRange angle_range
Angular range, in degrees, for random rotation.
String inpainter_path
Inpainter file path.
float delta_scale
Maximum variation for defects scaling.
float stretch_contrast
Maximum variation for histogram stretching.
float shift_horizontal
Shift along x-axis.
float shift_vertical
Shift along y-axis.
SVL parameters that configure the Shallow Learning training, it wraps the sb_t_svl_dl_tiling_par stru...
SbRangeFlt scale
Scale to applied to the image before the processing.
int num_tiles
Number of horizontal and vertical tiles used to process the image.
SbTilingMode mode
Automatic tiling mode for image processing.
Svl parameters (not used at the moment) that wraps the sb_t_svl_par structure
SbImageCircularityType image_circularity_type
Image circularity type.
int reproducibility
Enable the reproducibility of the training.
String image_ext
Extensions of the images.
sb_fp_svl_progress fp_progress
The SVL calls this callback to notify the user the results of SVL.
void * user_data
Pointer to data which is passed to the callbacks.
SbSvlDlPar dl
Deep Learning modules parameters.
SbDevicesPar devices
Devices used for training.
int num_threads
Maximum number of OpenMP threads that SVL can use.
SbSvlSlPar sl
Shallow Learning modules parameters.
sb_fp_svl_command fp_command
Callback called by SVL to allow the user to decide how to continue when particular situations happen,...
String project_path
Path of the project, where the SVL will find the images.
float free_memory_percentage
Percentage of system memory that the svl tries to leave free.
SVL parameters that configure the Shallow Learning training, it wraps the sb_t_svl_sl_par structure
String features
List of the features among which SVL will choose the best features.
float goodness_target
Goodness target of the training.
SbSvlParOptimizationMode optimization_mode
Optimization mode.
int auto_levels
Enable the automatic surface levels training.
SbInterpolationMode
Interpolation mode enumeration that wraps the sb_t_interpolation_mode enum
SbImageFormat
Image format enumeration that wraps the sb_t_image_format enum
SbProjectType
Project type that wraps the sb_t_project_type enum
SbPerturbationType
Enumerates the range of application of the perturbations, it wraps the sb_t_perturbation_type enums.
SbImageBordersExtensionMode
Enumerates the image borders extension mode, it wraps the sb_t_image_borders_extension_mode enum.
SbNetworkFreezeMode
Deep learning network freeze mode that wraps the sb_t_network_freeze_mode enum
SbLossFnType
Enumerates the type of loss function, it wraps the sb_t_loss_fn_type enum.
SbSvlParOptimizationMode
Svl par optimization mode that wraps the sb_t_svl_par_optimization_mode enum
SbNetworkType
Deep Learning Network type that wraps the sb_t_network_type enum
SbFloatingPointPrecisionType
Floating point precision type that wraps the sb_t_floating_point_op_type enum
SbPerturbationMode
Enumerates the mode of the perturbations, it wraps the sb_t_perturbation_mode enum.
SbImageCircularityType
Image circularity that wraps the sb_t_image_circularity_type enum
SbTilingMode
Enumerates the tiling mode, it wraps the sb_t_tiling_mode enum.
@ SB_PROJECT_TYPE_RETINA
Project Retina.
@ SB_PROJECT_TYPE_DEEP_CORTEX
Project Deep Cortex.
@ SB_PROJECT_TYPE_DEEP_RETINA
Project Deep Retina.
@ SB_PROJECT_TYPE_SURFACE
Project Surface.
@ SB_PROJECT_TYPE_DEEP_SURFACE
Project Deep Surface.
@ SB_PERTURBATION_TYPE_IMAGE
The whole image is perturbated.
@ SB_PERTURBATION_TYPE_BOTH
Both image and defects perturbation types.
@ SB_PERTURBATION_TYPE_DEFECTS
Only the defect area is perturbated.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_KERNEL_AVG
Local average extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_BORDER_AVG
Last line average extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_V_127
Grey 127 extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_BORDER_AVG_LOCAL
Last line local average extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_V_0
Zero extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_V_255
Grey 255 extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_KERNEL_AVG_17x1
Horizontal local average extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_LAST
Last pixel extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_MIRRORING
Mirroring extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_NONE
Undefined extension mode.
@ SB_IMAGE_BORDERS_EXTENSION_MODE_IMAGE_AVG
Image average extension mode.
@ SB_NETWORK_FREEZE_MODE_ENCODER
Freeze the network encoder.
@ SB_NETWORK_FREEZE_MODE_NONE
Freeze mode disabled.
@ SB_NETWORK_FREEZE_MODE_FIRST_BLOCK
Freeze the first block of the newtork.
@ SB_NETWORK_FREEZE_MODE_ENCODER_DECODER
Freeze both encoder and decoder.
@ SB_LOSS_FN_TYPE_CCE
CCE loss.
@ SB_LOSS_FN_TYPE_FOCAL
Foacal loss.
@ SB_LOSS_FN_TYPE_BCE
BCE loss.
@ SB_SVL_PAR_OPTIMIZATION_USE_SELECTED
@ SB_SVL_PAR_OPTIMIZATION_TIME_FAST
@ SB_SVL_PAR_OPTIMIZATION_TIME_SLOW
@ SB_SVL_PAR_OPTIMIZATION_TIME_MEDIUM
@ SB_NETWORK_TYPE_ODNET0
Deep Learning Object Detection Network 0 with input size 416x416.
@ SB_NETWORK_TYPE_EFFICIENTNET_B1
Deep Learning EfficientNet b1.
@ SB_NETWORK_TYPE_SDINET0
Deep Learning Surface Defects Inspection Network 0 with variable input size.
@ SB_NETWORK_TYPE_ICNET0_128
Deep Learning Image Classification Network 0 with input size 128x128.
@ SB_NETWORK_TYPE_ICNET0_64
Deep Learning Image Classification Network 0 with input size 64x64.
@ SB_NETWORK_TYPE_EFFICIENTNET_B0
Deep Learning EfficientNet b0.
@ SB_NETWORK_TYPE_EFFICIENTNET_B2
Deep Learning EfficientNet b2.
@ SB_FLOATING_POINT_OPS_TYPE_HALF_PRECISION
Single Precision (Float16)
@ SB_FLOATING_POINT_OPS_TYPE_SINGLE_PRECISION
Single Precision (Float32)
@ SB_PERTURBATION_MODE_OFFLINE
Offline perturbation.
@ SB_PERTURBATION_MODE_ONLINE
Online perturbation.
@ SB_PERTURBATION_MODE_BOTH
Both online and offline perturbations.
@ SB_IMAGE_CIRCULARITY_TYPE_NONE
No image circularity.
@ SB_IMAGE_CIRCULARITY_TYPE_HORIZONTAL
Horizontal image circularity.
@ SB_IMAGE_CIRCULARITY_TYPE_VERTICAL
Vertical image circularity.
@ SB_TILING_MODE_AUTO
Auto-tiling.
@ SB_TILING_MODE_MANUAL
Manual tiling.
SbFrameworkType
Framework type that wraps the sb_t_framework_type enum
SbDeviceType
Device type that wraps the sb_t_device_type enum
sb_t_err(* sb_fp_svl_pre_elaboration)(const SB_HANDLE image_info, void *const user_data, const char *const image_file, int image_index, sb_t_svl_pre_elaboration *const pre_elaboration)
Definition of the callback for image pre elaboration.
sb_t_err(* sb_fp_svl_progress)(void *const user_data, sb_t_svl_res *res, int force)
Definition of the callback for the working progress.
sb_t_err(* sb_fp_svl_command)(void *const user_data, sb_t_svl_res *res, sb_t_svl_command *const command, const char *const msg)
Definition of the callback that the SVL calls when it needs to know how to continue.
Property of computational devices.
Describes the perturbation of a sample.
Shallow Learning modules parameters.
Deep Learning network parameters.
Describes the perturbation of the image / defect.
SVL parameters to configure the Deep Learning training.
SVL parameters that configure the Shallow Learning training.