jQfdddlmZddlZddlZddlZddlZddlZddlZddl Z ddl m cm Z ddlmZGddejZdKdLd Z dMdNdZdOdZdZdZdZd ZdPdQd%ZdRdSd*Zd+Zd,Zd-Zd.Zd/Zd0Z d1Z!dTdUd4Z"dVd7Z#dWdXd9Z$d:Z% dYdZd=Z&d[d\d?Z'd@Z(d]d^dFZ)d_dHZ*dIZ+dJZ,dS)`) annotationsN) NOT_MACOS14c4eZdZdZd ddZd Zd Zd Zd ZdS)Profilea7Ultralytics Profile class for timing code execution. Use as a decorator with @Profile() or as a context manager with 'with Profile():'. Provides accurate timing measurements with CUDA synchronization support for GPU operations. Attributes: t (float): Accumulated time in seconds. device (torch.device): Device used for model inference. cuda (bool): Whether CUDA is being used for timing synchronization. Examples: Use as a context manager to time code execution >>> with Profile(device=device) as dt: ... pass # slow operation here >>> print(dt) # prints "Elapsed time is 9.5367431640625e-07 s" Use as a decorator to time function execution >>> @Profile() ... def slow_function(): ... time.sleep(0.1) Ntfloatdevicetorch.device | Nonec||_||_t|o!t|d|_dS)zInitialize the Profile class. Args: t (float): Initial accumulated time in seconds. device (torch.device, optional): Device used for model inference to enable CUDA synchronization. cudaN)rr boolstr startswithr )selfrr s X/var/www/html/Carbon-Document/venv/lib/python3.11/site-packages/ultralytics/utils/ops.py__init__zProfile.__init__)s> BCKK$:$:6$B$BCC c8||_|S)z Start timing.)timestartrs r __enter__zProfile.__enter__4sYY[[  rcr||jz |_|xj|jz c_dS)z Stop timing.N)rrdtr)rtypevalue tracebacks r__exit__zProfile.__exit__9s.))++ * $'rcd|jdS)zIReturn a human-readable string representing the accumulated elapsed time.zElapsed time is z s)rrs r__str__zProfile.__str__>s,$&,,,,rc~|jr$tj|jt jS)z9Get current time with CUDA synchronization if applicable.)r torch synchronizer r perf_counterrs rrz Profile.timeBs3 9 0 J " "4; / / / """r)rN)rr r r ) __name__ __module__ __qualname____doc__rrrr!rrrrrsx, D D D D D  ---#####rrsegment np.ndarraywidthintheightreturnc|j\}}tj|dk|dk||k||kgdkr,|d|}|d|}|dk|dkz||kz||kz}||}||}t|rgtj||||g|jntj d|jS)a$Convert segment coordinates to bounding box coordinates. Converts a single segment label to a box label by finding the minimum and maximum x and y coordinates. Applies inside-image constraint and clips coordinates when necessary. Args: segment (np.ndarray): Segment coordinates in format (N, 2) where N is number of points. width (int): Width of the image in pixels. height (int): Height of the image in pixels. Returns: (np.ndarray): Bounding box coordinates in xyxy format [x1, y1, x2, y2]. rdtype) Tnparrayminmaxsumclipanyr5zeros)r,r.r0xyinsides r segment2boxrCIs/ 9DAq x1aeeggk15577U?AEEGGf|t|d|dz |d|dz }t|dt|d|zz dz dz }t|dt|d|zz dz dz }n|dd}|d\}}|rB|dxx|zcc<|dxx|zcc<|s |dxx|zcc<|d xx|zcc<|d dd fxx|zcc<|r|nt||S) aVRescale bounding boxes from one image shape to another. Rescales bounding boxes from img1_shape to img0_shape, accounting for padding and aspect ratio changes. Supports both xyxy and xywh box formats. Args: img1_shape (tuple[int, int]): Shape of the source image (height, width). boxes (torch.Tensor | np.ndarray): Bounding boxes to rescale in format (N, 4). img0_shape (tuple[int, int]): Shape of the target image (height, width). ratio_pad (tuple, optional): Tuple of (ratio, pad) for scaling. If None, calculated from image shapes. padding (bool): Whether boxes are based on YOLO-style augmented images with padding. xywh (bool): Whether box format is xywh (True) or xyxy (False). Returns: (torch.Tensor | np.ndarray): Rescaled bounding boxes in the same format as input. Nr皙?.r.rN.rO.r3.r6)r:round clip_boxes) rDrFrHrIrKrLgainpad_xpad_ys r scale_boxesrZfsP0:a=:a=0*Q-*Q-2OPPz!}uZ]T-A'B'BBaG#MNNz!}uZ]T-A'B'BBaG#MNN|A | u# f   f   # &MMMU "MMM &MMMU "MMM #rr'NNNdNNN ;55j ;;;rr@ct|tjr!t|}t j||z |zS)zReturn the nearest number that is divisible by the given divisor. Args: x (int): The number to make divisible. divisor (int | torch.Tensor): The divisor. Returns: (int): The nearest number divisible by the divisor. ) isinstancer#Tensorr/r;mathceil)r@divisors rmake_divisiblerasF'5<((%gkkmm$$ 9Q[ ! !G ++rc|dd\}}t|tjrtrq|dd||dd||dd||dd|n|dd||d<|dd||d<|dd||d<|dd||d<nN|dddgfd||dddgf<|dd d gfd||dd d gf<|S) aClip bounding boxes to image boundaries. Args: boxes (torch.Tensor | np.ndarray): Bounding boxes to clip. shape (tuple): Image shape as HWC or HW (supports both). Returns: (torch.Tensor | np.ndarray): Clipped bounding boxes. NrOrQrrRrSrT.rNr3r\r#r]rclamp_clampr=)rFshapehws rrVrVsu !9DAq%&& ;  6 &M A & & & &M A & & & &M A & & & &M A & & & &!&M//155E&M!&M//155E&M!&M//155E&M!&M//155E&MM"3A;/44Q::cAq6k"3A;/44Q::cAq6k Lrc|dd\}}t|tjrtr9|dd||dd|n}|dd||d<|dd||d<n>|dd||d<|dd||d<|S)a Clip line coordinates to image boundaries. Args: coords (torch.Tensor | np.ndarray): Line coordinates to clip. shape (tuple): Image shape as HWC or HW (supports both). Returns: (torch.Tensor | np.ndarray): Clipped coordinates. NrOrQrrRrc)coordsrfrgrhs r clip_coordsrks !9DAq&%,'' 3  8 6N ! !!Q ' ' ' 6N ! !!Q ' ' ' '#F^11!Q77F6N#F^11!Q77F6NN,,Q22v,,Q22v Mrc|jddksJd|jt|}|d|d|d|df\}}}}||zdz |d<||zdz |d<||z |d<||z |d<|S) aConvert bounding box coordinates from (x1, y1, x2, y2) format to (x, y, width, height) format where (x1, y1) is the top-left corner and (x2, y2) is the bottom-right corner. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates in (x1, y1, x2, y2) format. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in (x, y, width, height) format. r69input shape last dimension expected 4 but input shape is rQrRrSrTrOrf empty_like)r@rAx1y1x2y2s r xyxy2xywhrus 72;!   bYZY`bb   1 AvY& 1V9ai?NBBbA AfIbA AfIRAfIRAfI Hrc|jddksJd|jt|}|dddf}|dddfdz }||z |dddf<||z|dddf<|S)aConvert bounding box coordinates from (x, y, width, height) format to (x1, y1, x2, y2) format where (x1, y1) is the top-left corner and (x2, y2) is the bottom-right corner. Note: ops per 2 channels faster than per channel. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates in (x, y, width, height) format. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in (x1, y1, x2, y2) format. rmr6rn.NrOro)r@rAxywhs r xywh2xyxyrys 72;!   bYZY`bb   1 A 37B 37aBbAc2A2gJbAc122gJ Hrrhrgpadwpadhc4|jddksJd|jt|}|d|d|d|df\}}}} |dz | dz } } ||| z z|z|d<||| z z|z|d<||| zz|z|d<||| zz|z|d<|S) aConvert normalized bounding box coordinates to pixel coordinates. Args: x (np.ndarray | torch.Tensor): Normalized bounding box coordinates in (x, y, w, h) format. w (int): Image width in pixels. h (int): Image height in pixels. padw (int): Padding width in pixels. padh (int): Padding height in pixels. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in (x1, y1, x2, y2) format. rmr6rnrQrRrSrTrOro) r@rhrgrzr{rAxcycxwxhhalf_whalf_hs r xywhn2xyxyrs 72;!   bYZY`bb   1 AvY& 1V9ai?NBB!VR!VFFR&[!D(AfIR&[!D(AfIR&[!D(AfIR&[!D(AfI Hrrr=epsr cH|rt|||z ||z f}|jddksJd|jt|}|d|d|d|df\}}}} ||zdz |z |d<|| zdz |z |d<||z |z |d<| |z |z |d<|S) a`Convert bounding box coordinates from (x1, y1, x2, y2) format to normalized (x, y, width, height) format. x, y, width and height are normalized to image dimensions. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates in (x1, y1, x2, y2) format. w (int): Image width in pixels. h (int): Image height in pixels. clip (bool): Whether to clip boxes to image boundaries. eps (float): Minimum value for box width and height. Returns: (np.ndarray | torch.Tensor): Normalized bounding box coordinates in (x, y, width, height) format. rmr6rnrQrRrSrTrO)rVrfrp) r@rhrgr=rrArqrrrsrts r xyxy2xywhnrs . q1s7AG, - - 72;!   bYZY`bb   1 AvY& 1V9ai?NBBr'Q!#AfIr'Q!#AfIbA AfIbA AfI Hrct|tjr|nt j|}|d|ddz z |d<|d|ddz z |d<|S)a(Convert bounding box format from [x, y, w, h] to [x1, y1, w, h] where x1, y1 are top-left coordinates. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates in xywh format. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in ltwh format. rQrSrOrRrTr\r#r]cloner8copyr@rAs r xywh2ltwhr,g 5<00@ bgajjA& AfIM)AfI& AfIM)AfI Hrct|tjr|nt j|}|d|dz |d<|d|dz |d<|S)aConvert bounding boxes from [x1, y1, x2, y2] to [x1, y1, w, h] format. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates in xyxy format. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in ltwh format. rSrQrTrRrrs r xyxy2ltwhr;_ 5<00@ bgajjA& AfI%AfI& AfI%AfI Hrct|tjr|nt j|}|d|ddz z|d<|d|ddz z|d<|S)a Convert bounding boxes from [x1, y1, w, h] to [x, y, w, h] where xy1=top-left, xy=center. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in xywh format. rQrSrOrRrTrrs r ltwh2xywhrJrrct|tj}|r&|n|}|t |dd}g}|D]}tj|\\}}\}}} | dz tj z} ||kr||}}| tj dz z } | dtj zdz kr%| tj z} | dtj zdz k%| tj dz kr#| tj z } | tj dz k#| ||||| g|r!tj ||j |jntj|S)aConvert batched Oriented Bounding Boxes (OBB) from [xy1, xy2, xy3, xy4] to [xywh, rotation] format. Args: x (np.ndarray | torch.Tensor): Input box corners with shape (N, 8) in [xy1, xy2, xy3, xy4] format. Returns: (np.ndarray | torch.Tensor): Converted data in [cx, cy, w, h, rotation] format with shape (N, 5). Rotation values are in radians from [-pi/4, 3pi/4). rmrOr3r6r r5)r\r#r]cpunumpyreshapelencv2 minAreaRectr8piappendtensorr r5asarray) r@is_torchpointsrboxesptscxcyrhrganglethetas rxyxyxyxy2xywhrrYsp!U\**H ( /QUUWW]]___aF ^^CFFB * *F F - -#&/#"6"6R&1a% be# q55aqA RUQY Eq25y1}$$ RUNEq25y1}$$rufqj  RUNErufqj   r2q!U+,,,,CK c5<qxqw ? ? ? ?QSQ[\bQcQccrcPttjr.tjtjtjtjfn-tjtjtjtjf\}}}}dddf}fdtddD\}}}||||} } |dz | z|dz | zg} | dz | z|dz | zg} || d} || d} || z| z} || z| z }|| z | z }|| z | z}|| |||gdS)aConvert batched Oriented Bounding Boxes (OBB) from [xywh, rotation] to [xy1, xy2, xy3, xy4] format. Args: x (np.ndarray | torch.Tensor): Boxes in [cx, cy, w, h, rotation] format with shape (N, 5) or (B, N, 5). Rotation values should be in radians from [-pi/4, 3pi/4). Returns: (np.ndarray | torch.Tensor): Converted corner points with shape (N, 4, 2) or (B, N, 4, 2). .NrOc36K|]}d||dzfVdS).rNNr*).0ir@s r z!xywhr2xyxyxyxy..s4::1S!a!e)^$::::::rrm) r\r#r]cossincatstackr8 concatenaterange)r@rrrrctrrhrgr cos_value sin_valuevec1vec2pt1pt2pt3pt4s` rxywhr2xyxyxyxyrxsN a & & 8EIuy%+66fbfbnbh 7Cc5 C!G*C::::eAqkk:::KAq%3u::ss5zzyI EI q1uy0 1D BFY A 1 2D 3tR==D 3tR==D *t C *t C *t C *t C 5#sC%r * **rct|tjr|nt j|}|d|dz|d<|d|dz|d<|S)aConvert bounding box from [x1, y1, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right. Args: x (np.ndarray | torch.Tensor): Input bounding box coordinates. Returns: (np.ndarray | torch.Tensor): Bounding box coordinates in xyxy format. rSrQrTrRrrs r ltwh2xyxyrrrc(g}|D]m}|j\}}|||||gnt t j|S)aConvert segment coordinates to bounding box labels in xywh format. Args: segments (list): List of segments where each segment is a list of points, each point is [x, y] coordinates. Returns: (np.ndarray): Bounding box coordinates in xywh format. )r7rr:r;rur8r9)segmentsrFsr@rAs rsegments2boxesrsw E ;;s1 aeeggquuww9:::: RXe__ % %%rnc t|D]?\}t|krtjddddffdtjdtdz t|kr|tz n|tjtt|kr)tjtjntjfdtdDtj  ddj ||<A|S) a2Resample segments to n points each using linear interpolation. Args: segments (list): List of (N, 2) arrays where N is the number of points in each segment. n (int): Number of points to resample each segment to. Returns: (list): Resampled segments with n points each. rrNN)axisc Pg|]"}tjdd|f#S)N)r8interp)rrrr@xps r z%resample_segments..s3HHH!BIaQqqq!tW55HHHrrOr4rm) enumeraterr8rlinspacearangeinsert searchsortedrfloat32rr7)rrrrr@rs @@@rresample_segmentsrs:(##   1 q66Q;;  NAq1aaay> 2 2 2 K3q66A:SVVaZZq3q66zzQ G G Ys1vv  8;A BIaB// 4 4 4 NHHHHHHuQxxHHHPRPZ [ [ [ c cdegi j j l   Ormasks torch.Tensorc|j|jkr||j}|j\}}}|dkr|js~t |dD]4\}\}}}} d||d|f<d||| df<d||ddd|f<d||dd|df<5|Stj |dddddfdd\}}}} tj ||j|j ddddf} tj ||j|j ddddf} || |k| |kz| |kz| | kzzS)aCrop masks to bounding box regions. Args: masks (torch.Tensor): Masks with shape (N, H, W). boxes (torch.Tensor): Bounding box coordinates with shape (N, 4) in xyxy pixel format. Returns: (torch.Tensor): Cropped masks. 2r)r:Nr6rNr) r torfis_cudarrerUr/r#chunkrr5) rrFrrgrhrrqrrrsrtrcs r crop_maskrs |u|##&&kGAq!2vvemv#,U[[Q[-?-?-E-E-G-G-K-K-M-M#N#N ! ! ABBE!SbS&ME!RSS&M E!QQQ)  E!QQQ)   U111aaa:%61==BB L5>    rc>|j\}}}|||dzd||}t|d|d}t ||}|dS)aApply masks to bounding boxes using mask head output with native upsampling. Args: protos (torch.Tensor): Mask prototypes with shape (mask_dim, mask_h, mask_w). masks_in (torch.Tensor): Mask coefficients with shape (N, mask_dim) where N is number of masks after NMS. bboxes (torch.Tensor): Bounding boxes with shape (N, 4) where N is number of masks after NMS. shape (tuple): Input image size as (height, width). Returns: (torch.Tensor): Binary mask tensor with shape (N, H, W). rmNrr)rfr r scale_masksrrr)rrrrfrrrrs rprocess_mask_nativers IAr2  ++Ar22 2 8 8R D DE d U + +A .E eV $ $E 99S>>    rrf.tuple[tuple[int, int], tuple[int, int]] | Nonec&|jdd\}}|dd\}}||kr||kr|S|Mt||z ||z }|t||zz |t||zz } } |r | dz} | dz} n |d\} } |r$t| dz t| dz fnd\} } |t| dzz } |t| dzz }tj|d| | | |f|dS) aRescale segment masks to target shape. Args: masks (torch.Tensor): Masks with shape (N, C, H, W). shape (tuple[int, int]): Target height and width as (height, width). ratio_pad (tuple, optional): Ratio and padding values as ((ratio_h, ratio_w), (pad_w, pad_h)). padding (bool): Whether masks are based on YOLO-style augmented images with padding. Returns: (torch.Tensor): Rescaled masks. rONrNrP)rr.rr)rfr:rUrrr )rrfrIrKim1_him1_wim0_him0_wrWpad_wpad_htopleftbottomrights rrrsB";qrr?LE5!9LE5 ~~%5.. 55=%%-00edl 3 33uuUT\?R?R7Ru   QJE QJE | u.s&7T7T7TQ "a8H8H7T7T7TrrmrOlargestc,g|]}t|Sr*)rrs rrz"masks2segments..s(;(;(;AQ(;(;(;r)rrOr)ultralytics.data.converterrr\r8ndarrayastyperrrascontiguousarrayr findContours RETR_EXTERNALCHAIN_APPROX_SIMPLErrrr9argmaxr?r)rrrrr@rs rmasks2segmentsr$ns?>>>>>%/rz%B%B bELL ! ! ! HXHXHZHZH`H`HbHbEH  !% ( ( - -  Q 133J K KA N !5  1vvzzN#6#67T7TRS7T7T7T#U#UVVV1b!,, Y&&HQrx(;(;(;(;(;<<CCEEFGGOOPRTUVV  A++,,,, Orbatchc|dddddzddS)a_Convert a batch of FP32 torch tensors to NumPy uint8 arrays, changing from BCHW to BHWC layout. Args: batch (torch.Tensor): Input tensor batch with shape (Batch, Channels, Height, Width) and dtype torch.float32. Returns: (np.ndarray): Output NumPy array batch with shape (Batch, Height, Width, Channels) and dtype uint8. rrOr3rN)permute contiguousrerrr)r%s rconvert_torch2numpy_batchr*sa MM!Q1 % % 0 0 2 2S 8 ? ?3 G G L L N N R R T T Z Z \ \\rc0tjdd|S)zClean a string by replacing special characters with '_' character. Args: s (str): A string needing special characters replaced. Returns: (str): A string with special characters replaced by an underscore _. u![|@#!¡·$€%&()=?¿^*;:,¨`><+]_)patternreplstring)resub)rs r clean_strr2s 6=CPQ R R RRrct|tjrtj||jnt j||jS)zKCreate empty torch.Tensor or np.ndarray with same shape and dtype as input.r4)r\r#r]rpr5r8)r@s rrprpsF1;Au|1L1L q5 AQW - - - -RTR_`aijipRqRqRqqr)r+r+)r,r-r.r/r0r/r1r-)NTF)rDrErFrGrHrErIrJrKrrLrr1rG)r@r/)r+r+rr)rhr/rgr/rzr/r{r/)r+r+Fr)rhr/rgr/r=rrr )r)rr/)rrrFrr1r)F)rr)NT) rrrfrErIrrKrr1r)NFT)rrrKr)r)rrrrr1r)r%rr1r-)- __future__r contextlibr^r0rrrr8r#torch.nn.functionalnn functionalrultralytics.utilsrContextDecoratorrrCrZrarVrkruryrrrrrrrrrrrrrrrrr$r*r2rpr*rrr;s#"""""  ))))))4#4#4#4#4#j)4#4#4#nB# '<'<'<'<'+++<     & & & .EEEE6!!!!!6!!!,AE !]!]!]!]!]H     F222$> ] ] ] ] S S Srrrrrr