OpenCV
4.4.0
Open Source Computer Vision
|
The TrackerParams struct stores parameters of TrackerByMatching. More...
#include <opencv2/tracking/tracking_by_matching.hpp>
Public Member Functions | |
TrackerParams () | |
Public Attributes | |
float | aff_thr_fast |
float | aff_thr_strong |
cv::Vec2f | bbox_aspect_ratios_range |
Bounding box aspect ratios range. More... | |
cv::Vec2f | bbox_heights_range |
Bounding box heights range. More... | |
bool | drop_forgotten_tracks |
size_t | forget_delay |
int | max_num_objects_in_track |
float | min_det_conf |
Min confidence of detection. More... | |
size_t | min_track_duration |
Min track duration in milliseconds. More... | |
float | motion_affinity_w |
Motion affinity weight. More... | |
int | predict |
float | reid_thr |
Affinity threshold for re-identification. More... | |
float | shape_affinity_w |
Shape affinity weight. More... | |
float | strong_affinity_thr |
of lost track. More... | |
float | time_affinity_w |
Time affinity weight. More... | |
The TrackerParams struct stores parameters of TrackerByMatching.
cv::tbm::TrackerParams::TrackerParams | ( | ) |
Default constructor.
float cv::tbm::TrackerParams::aff_thr_fast |
Affinity threshold which is used to determine if tracklet and detection should be combined (fast descriptor is used).
float cv::tbm::TrackerParams::aff_thr_strong |
Affinity threshold which is used to determine if tracklet and detection should be combined(strong descriptor is used).
cv::Vec2f cv::tbm::TrackerParams::bbox_aspect_ratios_range |
Bounding box aspect ratios range.
cv::Vec2f cv::tbm::TrackerParams::bbox_heights_range |
Bounding box heights range.
bool cv::tbm::TrackerParams::drop_forgotten_tracks |
Drop forgotten tracks. If it's enabled it disables an ability to get detection log.
size_t cv::tbm::TrackerParams::forget_delay |
Forget about track if the last bounding box in track was detected more than specified number of frames ago.
int cv::tbm::TrackerParams::max_num_objects_in_track |
The number of objects in track is restricted by this parameter. If it is negative or zero, the max number of objects in track is not restricted.
float cv::tbm::TrackerParams::min_det_conf |
Min confidence of detection.
size_t cv::tbm::TrackerParams::min_track_duration |
Min track duration in milliseconds.
float cv::tbm::TrackerParams::motion_affinity_w |
Motion affinity weight.
int cv::tbm::TrackerParams::predict |
How many frames are used to predict bounding box in case
float cv::tbm::TrackerParams::reid_thr |
Affinity threshold for re-identification.
float cv::tbm::TrackerParams::shape_affinity_w |
Shape affinity weight.
float cv::tbm::TrackerParams::strong_affinity_thr |
of lost track.
If 'fast' confidence is greater than this threshold then 'strong' Re-ID approach is used.
float cv::tbm::TrackerParams::time_affinity_w |
Time affinity weight.