o oi:@sddlmZddlZddlZddlZddlmZddlmZm Z m Z ddl m Z ddl mZmZddlmZddlmZdd lmZdd lmZdd lmZdd lmZd gZeeZGdd d Z dS)) annotationsN)Path)AnyOptionalUnion) kornia_config)Tensortensor)boxmot)numpy) write_image)ObjectDetector)RTDETRDetectorBuilder)tensor_to_image BoxMotTrackercs^eZdZUdZdZded<     d$d%fdd Zd&ddZd'd(ddZd)d*d"d#Z Z S)+ra# BoxMotTracker is a module that wraps a detector and a tracker model. This module uses BoxMot library for tracking. Args: detector: ObjectDetector: The detector model. tracker_model_name: The name of the tracker model. Valid options are: - "BoTSORT" - "DeepOCSORT" - "OCSORT" - "HybridSORT" - "ByteTrack" - "StrongSORT" - "ImprAssoc" tracker_model_weights: Path to the model weights for ReID (Re-Identification). device: Device on which to run the model (e.g., 'cpu' or 'cuda'). fp16: Whether to use half-precision (fp16) for faster inference on compatible devices. per_class: Whether to perform per-class tracking track_high_thresh: High threshold for detection confidence. Detections above this threshold are used in the first association round. track_low_thresh: Low threshold for detection confidence. Detections below this threshold are ignored. new_track_thresh: Threshold for creating a new track. Detections above this threshold will be considered as potential new tracks. track_buffer: Number of frames to keep a track alive after it was last detected. match_thresh: Threshold for the matching step in data association. proximity_thresh: Threshold for IoU (Intersection over Union) distance in first-round association. appearance_thresh: Threshold for appearance embedding distance in the ReID module. cmc_method: Method for correcting camera motion. Options include "sof" (simple optical flow). frame_rate: Frame rate of the video being processed. Used to scale the track buffer size. fuse_first_associate: Whether to fuse appearance and motion information during the first association step. with_reid: Whether to use ReID (Re-Identification) features for association. .. code-block:: python import kornia image = kornia.utils.sample.get_sample_images()[0][None] model = BoxMotTracker() for i in range(4): # At least 4 frames are needed to initialize the tracking position model.update(image) model.save(image) .. note:: At least 4 frames are needed to initialize the tracking position. boxmot_trackerstrname rtdetr_r18vd DeepOCSORTosnet_x0_25_msmt17.ptcpuFdetectorUnion[ObjectDetector, str]tracker_model_nametracker_model_weightsdevicefp16boolkwargsrreturnNonec stt|tr|drtj|d}ntd|d||_t j t j dddt t|d tt jt j d|||d||_dS) Nrtdetr) model_namez Detector `zA` not available. You may pass an ObjectDetector instance instead.z/boxmotTexist_ok) model_weightsrr)super__init__ isinstancer startswithrbuild ValueErrorrosmakedirsrhub_models_dirgetattrr rpathjointracker)selfrrrrrr __class__r'Y/home/ubuntu/.local/lib/python3.10/site-packages/kornia/models/tracking/boxmot_tracker.pyr)Zs     zBoxMotTracker.__init__imagerc Cs|jdkr |jddks|jdkstd|j|jdkr#|d}||}|d}t|dddf|dddf|dddf|dddf|dddf|dddf|dddf|dddfgj }|jddkr|t d }t |d  tj }|j||S) zaUpdate the tracker with a new image. Args: image: The input image. rz=Input tensor must be of shape (1, 3, H, W) or (3, H, W). Got N)r)ndimshaper- unsqueezerrr nparrayTemptyrastypeuint8r4update)r5r9detections_raw detections frame_rawr'r'r8rJts*"     zBoxMotTracker.updateTshow_trajectoriescCs6t|dtj}|jj||dt|dddS)zVisualize the results of the tracker. Args: image: The input image. show_trajectories: Whether to show the trajectories. Returns: The image with the results of the tracker. r@rNr=rr;)rrHrDrIr4 plot_resultsr permute)r5r9rNrMr'r'r8 visualizes zBoxMotTracker.visualizeN directory Optional[str]cCs|dur|jdtjjtjjdd}tjd|}|j ||d}tj |ddt tj|t d  d d |td |dS) zSave the model to ONNX format. Args: image: The input image. show_trajectories: Whether to visualize trajectories. directory: Where to save the file(s). N_)tzz %Y%m%d%H%M%Skornia_outputsrOTr$rr?z.jpgzOutputs are saved in )rdatetimenowtimezoneutcstrftimer.r2r3rRr/r rzfillbyteloggerinfo)r5r9rNrSroutputr'r'r8saves $zBoxMotTracker.save)rrrrF)rrrrrrrrrrrrr r!)r9rr r!)T)r9rrNrr r)TN)r9rrNrrSrTr r!) __name__ __module__ __qualname____doc__r__annotations__r)rJrRrb __classcell__r'r'r6r8r(s  /  $)! __future__rrXloggingr.pathlibrtypingrrr kornia.configr kornia.corerr kornia.core.externalr r rD kornia.ior kornia.models.detection.baser kornia.models.detection.rtdetrrkornia.utils.imager__all__ getLoggerrcr_rr'r'r'r8s"