ObjectDetectionRuntimeParameters Class Reference

Class containing a set of runtime parameters for the object detection module. More...

Functions
ObjectDetectionRuntimeParameters	__cinit__ (self, detection_confidence_threshold=50, object_class_filter=[], object_class_detection_confidence_threshold={})
	Default constructor. More...
float	detection_confidence_threshold (self)
	Confidence threshold. More...
list[OBJECT_CLASS]	object_class_filter (self)
	Defines which object types to detect and track. More...
{}	object_class_detection_confidence_threshold (self)
	Dictonary of confidence thresholds for each class (can be empty for some classes). More...

Detailed Description

Class containing a set of runtime parameters for the object detection module.

The default constructor sets all parameters to their default settings.

Note

Parameters can be adjusted by the user.

Functions

__cinit__()

ObjectDetectionRuntimeParameters __cinit__	(		self,
			detection_confidence_threshold = 50,
			object_class_filter = [],
			object_class_detection_confidence_threshold = {}
	)

Default constructor.

All the parameters are set to their default values.

Parameters

detection_confidence_threshold	: Chosen detection_confidence_threshold
object_class_filter	: Chosen object_class_filter
object_class_detection_confidence_threshold	: Chosen object_class_detection_confidence_threshold

detection_confidence_threshold()

float detection_confidence_threshold

(

self

)

Confidence threshold.

From 1 to 100, with 1 meaning a low threshold, more uncertain objects and 99 very few but very precise objects.
Default: 20

Note

If the scene contains a lot of objects, increasing the confidence can slightly speed up the process, since every object instance is tracked.

detection_confidence_threshold is used as a fallback when sl::ObjectDetectionRuntimeParameters.object_class_detection_confidence_threshold is partially set.

object_class_filter()

list[OBJECT_CLASS] object_class_filter

(

self

)

Defines which object types to detect and track.

Default: [] (all classes are tracked)

Note

Fewer object types can slightly speed up the process since every object is tracked.

Will output only the selected classes.

In order to get all the available classes, the filter list must be empty :

object_class_filter = {};

To select a set of specific object classes, like vehicles, persons and animals for instance:

object_class_filter = {sl.OBJECT_CLASS.VEHICLE, sl.OBJECT_CLASS.PERSON, sl.OBJECT_CLASS.ANIMAL};

object_class_detection_confidence_threshold()

{} object_class_detection_confidence_threshold

(

self

)

Dictonary of confidence thresholds for each class (can be empty for some classes).

Note

sl.ObjectDetectionRuntimeParameters.detection_confidence_threshold will be taken as fallback/default value.