a yfRQ@sjddlZddlmZddlmZddlmZmZddlm Z ddl m Z Gd d d eZ Gd d d Z dS) N)LOGGER) xywh2ltwh) BaseTrack TrackState)matchingKalmanFilterXYAHcseZdZdZeZfddZddZeddZ ee dd fd d Z d d Z d#ddZddZddZeddZeddZeddZeddZeddZedd Zd!d"ZZS)$STracka Single object tracking representation that uses Kalman filtering for state estimation. This class is responsible for storing all the information regarding individual tracklets and performs state updates and predictions based on Kalman filter. Attributes: shared_kalman (KalmanFilterXYAH): Shared Kalman filter that is used across all STrack instances for prediction. _tlwh (np.ndarray): Private attribute to store top-left corner coordinates and width and height of bounding box. kalman_filter (KalmanFilterXYAH): Instance of Kalman filter used for this particular object track. mean (np.ndarray): Mean state estimate vector. covariance (np.ndarray): Covariance of state estimate. is_activated (bool): Boolean flag indicating if the track has been activated. score (float): Confidence score of the track. tracklet_len (int): Length of the tracklet. cls (Any): Class label for the object. idx (int): Index or identifier for the object. frame_id (int): Current frame ID. start_frame (int): Frame where the object was first detected. Methods: predict(): Predict the next state of the object using Kalman filter. multi_predict(stracks): Predict the next states for multiple tracks. multi_gmc(stracks, H): Update multiple track states using a homography matrix. activate(kalman_filter, frame_id): Activate a new tracklet. re_activate(new_track, frame_id, new_id): Reactivate a previously lost tracklet. update(new_track, frame_id): Update the state of a matched track. convert_coords(tlwh): Convert bounding box to x-y-aspect-height format. tlwh_to_xyah(tlwh): Convert tlwh bounding box to xyah format. Examples: Initialize and activate a new track >>> track = STrack(xywh=[100, 200, 50, 80, 0], score=0.9, cls="person") >>> track.activate(kalman_filter=KalmanFilterXYAH(), frame_id=1) cstt|dvs(Jdt|tjt|ddtjd|_d|_d\|_ |_ d|_ ||_ d|_ ||_|d |_t|d kr|dnd|_dS) aX Initialize a new STrack instance. Args: xywh (List[float]): Bounding box coordinates and dimensions in the format (x, y, w, h, [a], idx), where (x, y) is the center, (w, h) are width and height, [a] is optional aspect ratio, and idx is the id. score (float): Confidence score of the detection. cls (Any): Class label for the detected object. Examples: >>> xywh = [100.0, 150.0, 50.0, 75.0, 1] >>> score = 0.9 >>> cls = "person" >>> track = STrack(xywh, score, cls) >zexpected 5 or 6 values but got NZdtype)NNFrr )super__init__lennpasarrayrfloat32_tlwh kalman_filtermean covariance is_activatedscore tracklet_lenclsidxangle)selfxywhrr __class__]/var/www/html/django/DPS/env/lib/python3.9/site-packages/ultralytics/trackers/byte_tracker.pyr3s   zSTrack.__init__cCs:|j}|jtjkrd|d<|j||j\|_|_dS)zTPredicts the next state (mean and covariance) of the object using the Kalman filter.rN)rcopystaterTrackedrpredictr)r!Z mean_stater%r%r&r+Qs  zSTrack.predictcCst|dkrdStdd|D}tdd|D}t|D] \}}|jtjkr@d||d<q@tj ||\}}tt ||D] \}\}}|||_ |||_ qdS)zgPerform multi-object predictive tracking using Kalman filter for the provided list of STrack instances.rNcSsg|]}|jqSr%rr(.0str%r%r& ]z(STrack.multi_predict..cSsg|] }|jqSr%rr-r%r%r&r0^r1r') rrr enumerater)rr*r shared_kalman multi_predictziprr)stracks multi_meanmulti_covarianceir/rcovr%r%r&r5Xs   zSTrack.multi_predictrc Cst|dkrtdd|D}tdd|D}|ddddf}ttjdtd|}|dddf}tt||D]R\}\}} ||}|dd|7<|| | } |||_ | ||_ q|dS) z\Update state tracks positions and covariances using a homography matrix for multiple tracks.rcSsg|]}|jqSr%r,r-r%r%r&r0kr1z$STrack.multi_gmc..cSsg|] }|jqSr%r2r-r%r%r&r0lr1Nrrr) rrrZkroneyefloatr3r6dotZ transposerr) r7Hr8r9RZR8x8tr:rr;r%r%r& multi_gmcgs   zSTrack.multi_gmccCsX||_||_|j||j\|_|_d|_t j |_ |dkrHd|_ ||_ ||_dS)zaActivate a new tracklet using the provided Kalman filter and initialize its state and covariance.rrTN)rnext_idtrack_idZinitiateconvert_coordsrrrrrr*r)rframe_id start_frame)r!rrGr%r%r&activatezs zSTrack.activateFcCsp|j|j|j||j\|_|_d|_tj|_ d|_ ||_ |rL| |_ |j|_|j|_|j|_|j|_dS)zbReactivates a previously lost track using new detection data and updates its state and attributes.rTN)rupdaterrrFtlwhrrr*r)rrGrDrErrr r)r! new_trackrGnew_idr%r%r& re_activates  zSTrack.re_activatecCsn||_|jd7_|j}|j|j|j||\|_|_tj |_ d|_ |j |_ |j |_ |j|_|j|_dS)a Update the state of a matched track. Args: new_track (STrack): The new track containing updated information. frame_id (int): The ID of the current frame. Examples: Update the state of a track with new detection information >>> track = STrack([100, 200, 50, 80, 0.9, 1]) >>> new_track = STrack([105, 205, 55, 85, 0.95, 1]) >>> track.update(new_track, 2) rTN)rGrrKrrJrrrFrr*r)rrrr r)r!rLrGZnew_tlwhr%r%r&rJs z STrack.updatecCs ||S)zZConvert a bounding box's top-left-width-height format to its x-y-aspect-height equivalent.) tlwh_to_xyah)r!rKr%r%r&rFszSTrack.convert_coordscCs^|jdur|jS|jdd}|d|d9<|dd|ddd8<|S)zYReturns the bounding box in top-left-width-height format from the current state estimate.Nrrr<)rrr(r!retr%r%r&rKs    z STrack.tlwhcCs*|j}|dd|dd7<|S)zjConverts bounding box from (top left x, top left y, width, height) to (min x, min y, max x, max y) format.rN)rKr(rPr%r%r&xyxys z STrack.xyxycCsFt|}|dd|ddd7<|d|d<|S)zWConvert bounding box from tlwh format to center-x-center-y-aspect-height (xyah) format.Nrr<)rrr()rKrQr%r%r&rOs zSTrack.tlwh_to_xyahcCs4t|j}|dd|ddd7<|S)z_Returns the current position of the bounding box in (center x, center y, width, height) format.Nr)rrrKr(rPr%r%r&r"s z STrack.xywhcCs0|jdurtd|jSt|j|jdgS)zcReturns position in (center x, center y, width, height, angle) format, warning if angle is missing.Nu@WARNING ⚠️ `angle` attr not found, returning `xywh` instead.)r rwarningr"r concatenater!r%r%r&xywhas  z STrack.xywhacCs2|jdur|jn|j}||j|j|j|jgS)zLReturns the current tracking results in the appropriate bounding box format.N)r rRrVtolistrErrr)r!Zcoordsr%r%r&resultsz STrack.resultcCsd|jd|jd|jdS)zdReturns a string representation of the STrack object including start frame, end frame, and track ID.ZOT_z_(-))rErH end_framerUr%r%r&__repr__szSTrack.__repr__)F)__name__ __module__ __qualname____doc__r r4rr+ staticmethodr5rr=rCrIrNrJrFpropertyrKrRrOr"rVrXr\ __classcell__r%r%r#r&r s2$         r c@s~eZdZdZdddZdddZdd Zdd d Zd d ZddZ e ddZ ddZ e ddZ e ddZe ddZdS) BYTETrackera; BYTETracker: A tracking algorithm built on top of YOLOv8 for object detection and tracking. Responsible for initializing, updating, and managing the tracks for detected objects in a video sequence. It maintains the state of tracked, lost, and removed tracks over frames, utilizes Kalman filtering for predicting the new object locations, and performs data association. Attributes: tracked_stracks (List[STrack]): List of successfully activated tracks. lost_stracks (List[STrack]): List of lost tracks. removed_stracks (List[STrack]): List of removed tracks. frame_id (int): The current frame ID. args (Namespace): Command-line arguments. max_time_lost (int): The maximum frames for a track to be considered as 'lost'. kalman_filter (KalmanFilterXYAH): Kalman Filter object. Methods: update(results, img=None): Updates object tracker with new detections. get_kalmanfilter(): Returns a Kalman filter object for tracking bounding boxes. init_track(dets, scores, cls, img=None): Initialize object tracking with detections. get_dists(tracks, detections): Calculates the distance between tracks and detections. multi_predict(tracks): Predicts the location of tracks. reset_id(): Resets the ID counter of STrack. joint_stracks(tlista, tlistb): Combines two lists of stracks. sub_stracks(tlista, tlistb): Filters out the stracks present in the second list from the first list. remove_duplicate_stracks(stracksa, stracksb): Removes duplicate stracks based on IoU. Examples: Initialize BYTETracker and update with detection results >>> tracker = BYTETracker(args, frame_rate=30) >>> results = yolo_model.detect(image) >>> tracked_objects = tracker.update(results) cCsHg|_g|_g|_d|_||_t|d|j|_||_ | dS)a Initialize a BYTETracker instance for object tracking. Args: args (Namespace): Command-line arguments containing tracking parameters. frame_rate (int): Frame rate of the video sequence. Examples: Initialize BYTETracker with command-line arguments and a frame rate of 30 >>> args = Namespace(track_buffer=30) >>> tracker = BYTETracker(args, frame_rate=30) rg>@N) tracked_stracks lost_stracksremoved_stracksrGargsintZ track_buffer max_time_lostget_kalmanfilterrreset_id)r!riZ frame_rater%r%r&rs  zBYTETracker.__init__Nc%sz|jd7_g}g}g}g}|j}t|dr4|jn|j}tj|tt| ddgdd}|j } ||j j k} ||j j k} ||j j k} | | @} || }|| }|| }|| }| | }| | }|||||g}g}|jD] }|js||q||q|||j|t|drT|durT|j||}t|t|||}tj||j jd\}}}|D]^\}}|}|}|jtjkr| ||j||n|j!||jdd ||q||||||}fd d |D} t"| |}tj|d d\}}}!|D]^\}}| |}||}|jtjkrb| ||j||n|j!||jdd ||q"|D].}"| |"}|jtj#kr|$||qfd d |D||}tj|dd\}}#}|D].\}}|| ||j|||q|#D] }"||"}|%||q |D]<}$|$}|j&|j j'krfqF|(|j)|j||qF|jD],}|j|j*|j+kr|%||qdd |jD|_||j||_||j||_|,|j|j|_|j-||,|j|j.|_|/|j|j\|_|_|j.-|t|j.dkr^|j.dd|_.tj0dd |jDtj1dS)zXUpdates the tracker with new detections and returns the current list of tracked objects.rxywhrr)ZaxisgmcN)ZthreshF)rMcs$g|]}|jtjkr|qSr%r)rr*r.r:) strack_poolr%r&r0_r1z&BYTETracker.update..g?csg|] }|qSr%r%rq) detectionsr%r&r0sr1gffffff?cSsg|]}|jtjkr|qSr%rpr.rBr%r%r&r0r1iicSsg|]}|jr|jqSr%)rrX)r.xr%r%r&r0r1r)2rGconfhasattrrnr"rrTZarangerZreshaperriZtrack_high_threshZtrack_low_thresh init_trackrfrappend joint_stracksrgr5roapplyr rC get_distsrZlinear_assignmentZ match_threshr)rr*rJrN iou_distanceZLostZ mark_lostZ mark_removedrZnew_track_threshrIrr[rk sub_stracksextendrhremove_duplicate_stracksrr)%r!resultsimgZactivated_stracksZrefind_stracksrgrhscoresZbboxesrZ remain_indsZinds_lowZ inds_highZ inds_secondZ dets_seconddetsZ scores_keepZ scores_secondZcls_keepZ cls_secondZ unconfirmedrftrackZwarpdistsmatchesZu_trackZ u_detectionZitrackedZidetZdetZdetections_secondZr_tracked_stracksZu_detection_seconditZ u_unconfirmedZinewr%)rsrrr&rJ%s$                    zBYTETracker.updatecCstS)zRReturns a Kalman filter object for tracking bounding boxes using KalmanFilterXYAH.r rUr%r%r&rlszBYTETracker.get_kalmanfiltercCs"t|rddt|||DSgS)zgInitializes object tracking with given detections, scores, and class labels using the STrack algorithm.cSsg|]\}}}t|||qSr%)r )r.rRscr%r%r&r0r1z*BYTETracker.init_track..)rr6)r!rrrrr%r%r&rxszBYTETracker.init_trackcCs$t||}|jjr t||}|S)z\Calculates the distance between tracks and detections using IoU and optionally fuses scores.)rr}riZ fuse_score)r!tracksrsrr%r%r&r|s  zBYTETracker.get_distscCst|dS)z@Predict the next states for multiple tracks using Kalman filter.N)r r5)r!rr%r%r&r5szBYTETracker.multi_predictcCs tdS)z_Resets the ID counter for STrack instances to ensure unique track IDs across tracking sessions.N)r rmr%r%r%r&rmszBYTETracker.reset_idcCs.g|_g|_g|_d|_||_|dS)zjResets the tracker by clearing all tracked, lost, and removed tracks and reinitializing the Kalman filter.rN)rfrgrhrGrlrrmrUr%r%r&resets  zBYTETracker.resetcCsXi}g}|D]}d||j<||q |D](}|j}||ds*d||<||q*|S)zcCombines two lists of STrack objects into a single list, ensuring no duplicates based on track IDs.rr)rEryget)tlistatlistbexistsresrBtidr%r%r&rzs    zBYTETracker.joint_strackscs dd|Dfdd|DS)zGFilters out the stracks present in the second list from the first list.cSsh|] }|jqSr%rErtr%r%r& r1z*BYTETracker.sub_stracks..csg|]}|jvr|qSr%rrtZ track_ids_br%r&r0r1z+BYTETracker.sub_stracks..r%)rrr%rr&r~szBYTETracker.sub_stracksc st||}t|dk}ggt|D]N\}}||j||j}||j||j}||krp|q,|q,fddt|D}fddt|D} || fS)zYRemoves duplicate stracks from two lists based on Intersection over Union (IoU) distance.g333333?csg|]\}}|vr|qSr%r%r.r:rB)dupar%r&r0r1z8BYTETracker.remove_duplicate_stracks..csg|]\}}|vr|qSr%r%r)dupbr%r&r0r1) rr}rwherer6rGrHryr3) ZstracksaZstracksbZpdistpairspqZtimepZtimeqZresaZresbr%)rrr&rs    z$BYTETracker.remove_duplicate_stracks)re)N)N)r]r^r_r`rrJrlrxr|r5rarmrrzr~rr%r%r%r&rds "  r     rd)numpyrutilsrZ utils.opsrZ basetrackrrrZutils.kalman_filterr r rdr%r%r%r&s    `