sb_t_svl_err Struct Reference

Contains information regarding the error of the SVL. More...

#include <sb.h>

Collaboration diagram for sb_t_svl_err:

Data Fields
char *	msg
	Error message.
sb_t_err	err
	Error code sb_t_err.
char	file_name [2][512]
	Name of the file of the image that generates the error.
float	scale [2]
	Scale of the samples.
int	num_samples
	Number of samples.
char	model_name [2][SB_PAR_STRING_LEN]
	Name of the model of the samples that generate the error.
int	sample_index [2]
	Index of the samples that are too close. Index starts from 0.
sb_t_point	centre [2]
	Centre of the samples.
float	dist
	Distance, in pixel, between the samples.
float	min_dist
	Minimum allowed distance between samples.
int	models_number [2]
	Models number.
sb_t_svl_dl_par_network	par_network [2]
	Network parameters.
sb_t_loss_fn_type	loss_fn [2]
	Loss Function.

Detailed Description

Contains information regarding the error of the SVL.

The following table shows for each possible error which structure parameters are set by the sb_svl_run function:

Error code	Message	Parameters used	Project type
SB_ERR_CLS	SVL of a model is failed	msg err model_name[0]	Retina Surface
SB_ERR_DL_ROI_DEFECTS_NOT_SET	Not enough ROI defects area in SVL or validation image dataset for a model	msg err model_name[0]	Deep Surface
SB_ERR_DL_PRE_TRAINING_NETWORK_TYPE	The network type of pre-training file is different from that of the project	msg err par_network[]. type	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_FEATURES_MULTIPLIER	The feature multiplier parameter of pre-training file is different from that of the project	msg err par_network[]. features_multiplier	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_IMAGE_FORMAT	The image format of pre-training file is different from that of the project	msg err par_network[]. image_format	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_IMAGE_CHANNELS	The image number of channels of pre-training file is different from that of the project	msg err par_network[]. n_channels	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_INPUT_SIZE	The image input size of pre-training file is different from that of the project	msg err par_network[]. input_size	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_MODELS_NUMBER	The models number of pre-training file is different from that of the project	msg err models_number[]	Deep Cortex Deep Surface
SB_ERR_DL_PRE_TRAINING_LOSS_FN	The loss function of pre-training file is different from that of the project	msg err loss_fn[]	Deep Cortex Deep Surface
SB_ERR_IMAGE_SIZE_CIRCULAR	Image resolution is smaller than the minimum allowed to handle circularity, reduce the scale or increase image resolution	msg err file_name[0] par_network[0]	Deep Surface
SB_ERR_MODEL_NAME_UNDEFINED	A sample does not have the model name defined	msg err sample_index centre[0] file_name[0]	Retina Deep Cortex
SB_ERR_LEVELS_NUMBER	The difference between minimum and maximum scale of the samples is too big	msg err file_name[] model_name[] sample_index[] centre[] scale[]	Retina
SB_ERR_SAMPLE_BACKGROUND	A model has too few background samples to execute the training	msg err scale[0] model_name[0]	Surface
SB_ERR_SAMPLE_DISTANCE	Reduce value of parameters sb_t_par_model::obj_min_distance between samples	msg err file_name[0] model_name[] sample_index[] centre[] dist min_dist	Retina
SB_ERR_SAMPLE_NUMBER	A model has too few samples for learning	msg err num_samples model_name[0]	Retina Surface
SB_ERR_SAMPLE_WRONG	A sample has wrong data	msg err sample_index centre[0] model_name[0] file_name[0]	Retina Deep Cortex
All error codes beyond those documented in this table	Message get with sb_err_format function	err	Retina Surface Deep Surface Deep Cortex

Definition at line 14176 of file sb.h.

---

The documentation for this struct was generated from the following file:

• sb.h