FPh{TddlZddlZddlZddlmZddlZddlmZddl Z ddl Z ddl m Z mZmZddl mZddlmZmZmZmZmZddlmZmZmZddlmZGd d ZeZGd d Z eed eddfdZ!eddddddfdZ"ee jFde jHe jFde jJe jFde jHddddfdZ&ed!dZ'd"dZ(d#dZ)d$dZ*dedfd Z+y)%N)Path)Image ImageDraw ImageFont) __version__)LOGGER TryExceptops plt_settingsthreaded) check_font check_versionis_ascii)increment_pathc.eZdZdZdZddZedZy)Colorsa Ultralytics default color palette https://ultralytics.com/. This class provides methods to work with the Ultralytics color palette, including converting hex color codes to RGB values. Attributes: palette (list of tuple): List of RGB color values. n (int): The number of colors in the palette. pose_palette (np.array): A specific color palette array with dtype np.uint8. c`d}|Dcgc]}|jd|c}|_t|j|_t j gdgdgdgdgdgdgd gd gd gd gd gdgdgdgdgdgdgdgdgdgtj |_ycc}w)zEInitialize colors as hex = matplotlib.colors.TABLEAU_COLORS.values().)FF3838FF9D97FF701FFFB21DCFD23148F90A92CC173DDB861A933400D4BB2C99A800C2FF3445936473FF0018EC8438FF520085CB38FFFF95C8FF37C7#)r)r*3)r*f)r0r)r*r,r*)r,r*)r*r/r*)r*r-r*)r/r.r*)r-r,r*)r*r,r,)r*r/r/)r*r-r-)r,r*r,)r/r*r/)r-r*r-)rr*r)rrr*)r*rrr*r*r*dtypeN)hex2rgbpalettelennnparrayuint8 pose_palette)selfhexscs eC:\Users\daisl\Desktop\realtime-object-detection\venv\Lib\site-packages\ultralytics/utils/plotting.py__init__zColors.__init__"st7;z!Colors.hex2rgb..6s,D)QS1q5Q+R0)s"%)rrC)tuple)rNs`r@r5zColors.hex2rgb3sD)DDDrHN)F)__name__ __module__ __qualname____doc__rArG staticmethodr5rHr@rrs'  50 EErHrcLeZdZdZd dZd dZd dZddZddZddZ d Z d Z y) Annotatora7 Ultralytics Annotator for train/val mosaics and JPGs and predictions annotations. Attributes: im (Image.Image or numpy array): The image to annotate. pil (bool): Whether to use PIL or cv2 for drawing annotations. font (ImageFont.truetype or ImageFont.load_default): Font used for text annotations. lw (float): Line width for drawing. skeleton (List[List[int]]): Skeleton structure for keypoints. limb_color (List[int]): Color palette for limbs. kpt_color (List[int]): Color palette for keypoints. Nc||jjsJdt| }|xs|_|xs.t t t |jdz dzd_jrt|tjr|ntj|_ tjj_ t!|rdn|}|xs8t t t jj"dz dzd}t%j&t)||_t1t2drNfdj*_n9|_ t jd z d _jd z _d d gd dgd dgddgddgddgddgddgddgddgddgddgdd gd dgd d gddgd dgddgddgg_t<j>gd_ t<j>gd_!y#t,$rt%j._YwxYw)zjInitialize the Annotator class with image and line width along with color palette for keypoints and limbs.zQImage not contiguous. Apply np.ascontiguousarray(im) to Annotator() input images.rCg~jth?zArial.Unicode.ttfgQ? z9.2.0c@jj|ddS)NrCrP)fontgetbbox)xr=s r@z$Annotator.__init__..[sdii.?.?.B1Q.GrHr rK  rP)ririririrgrgrgrrrrrrKrKrKrKrKrKrK)rKrKrKrKrKrrrrrrririririririN)"data contiguousrpilmaxroundsumshapelw isinstancer fromarrayimrDrawdrawrsizertruetypestrr] Exception load_defaultr pil_versiongetsizetfsfskeletoncolorsr< limb_color kpt_color) r=rw line_width font_sizer]roexample non_asciirzs ` r@rAzAnnotator.__init__Jsww!!v#vv! )) #)HE#bhh-!*;e*C$Da H 88&r5;;7bU__R=PDG!tww/DI 5!"5M QCc$'',,.?!.Ce.K(Lb$Q%..s4y$? ['2$G !DG$''A+q)DGggkDGbB8b"XBx"bAr7UVXZT[^_ab]cfgijeknoqrmsR1b'Aq6Aq6Aq6Aq6Aq6TUWXSY\]_`[ac !--.no,,-ef 5%224  5sA/H"H;:H;c t|tjr|j}|js t |s|j j||j||r|jj|\}}|d|z dk\}|j j|d|r|d|z n|d|d|zdz|r|ddzn |d|zdzf||j j|d|r|d|z n|df|||jy y t|dt|dft|dt|df} }tj|j|| ||jtj|rtj |d|j"|j$ d\}}|d|z dk\}|d|z|r |d|z dz n |d|zdzf} tj|j|| |d tjtj&|j||d|r|ddz n |d|zdzfd|j"||j$tjy y ) z%Add one xyxy box to image with label.widthoutliner rfillrr]rCra thicknesslineType fontScalerN)rutorchTensortolistrorry rectanglertr]rtextrDcv2rwLINE_AA getTextSizerrputText) r=boxlabelcolor txt_colorwrNoutsidep1p2s r@ box_labelzAnnotator.box_labelgsI c5<< (**,C 888E? II  477E  Byy((/1a&1*/ ##V7SVaZAA Q#*SVaZA Q@$  AgA 3q6JEXahlhqhqr#a&k3s1v;/#c!f+s3q6{1KB MM$''2r5DGGckk Zua477dggVWXY1Q%!)q.UQY1 A beai!mK dggr2ub#++F DGG!BqE2a519RUQYQR]#S GG%&*gg%([[ 2 rHcV|jr2tj|jj |_t |dk(rM|j dddjjjdz|jdd|j|jk7r|j|j}tj||jtjdz }|ddddf}|jd}|||zz}d||zz j!d}|j#d j$}|j'dg }|j dddj}||d z|z}|dz} | j)jj} |r| n)t+j,| |jj.|jdd|jr|j1|jyy) a Plot masks on image. Args: masks (tensor): Predicted masks on cuda, shape: [n, h, w] colors (List[List[Int]]): Colors for predicted masks, [[r, g, b] * n] im_gpu (tensor): Image is in cuda, shape: [3, h, w], range: [0, 1] alpha (float): Mask transparency: 0.0 fully transparent, 1.0 opaque retina_masks (bool): Whether to use high resolution masks or not. Defaults to False. rr rCr*N)devicer4go@radim)dimsr)ror9asarrayrwcopyr7permuterncpunumpyrtortensorfloat32 unsqueezecumprodrpvaluesflipbyter scale_imagersrv) r=masksrim_gpualpha retina_masks masks_colorinv_alpha_masksmcsim_mask im_mask_nps r@rzAnnotator.maskss 88jj)..0DG u:?1a0;;=AACIIKcQDGGAJ ==ELL (YYu||,FfU\\ORWW4 &"v~. uu},55a8oo!o$++1#&1a(335/"--3C<\\^'')//1 #/ZS__ZQUQXQXQ^Q^5_ 88 NN477 # rHc ||jr2tj|jj |_|j \}}|dk(xr|dk(}||z}t |D]\}} |r&|j|D cgc] } t| c} n t|} | d| d} } | |dzdk7sO| |dzdk7s[t| dk(r | d}|dkrttj|jt| t| f|| dtj|ra|j d}t |jD]9\}}t||ddz dft||ddz dff}t||ddz dft||ddz dff}|dk(r%||ddz df}||ddz df}|dks|dkr|d|dzdk(s|d|dzdk(s|ddks|ddkr|d|dzdk(s|d|dzdk(s|ddks|ddkrtj|j|||j |D cgc] } t| c} dtj <|jr|j#|jy y cc} wcc} w) aQ Plot keypoints on the image. Args: kpts (tensor): Predicted keypoints with shape [17, 3]. Each keypoint has (x, y, confidence). shape (tuple): Image shape as a tuple (h, w), where h is the height and w is the width. radius (int, optional): Radius of the drawn keypoints. Default is 5. kpt_line (bool, optional): If True, the function will draw lines connecting keypoints for human pose. Default is True. Note: `kpt_line=True` currently only supports human pose plotting. rcrarr rC?r)rrN)ror9rrwrrs enumeraterrDrr7rcirclerrlinerrv)r=kptsrsradiuskpt_linenkptndimis_poserEkr_color_kx_coordy_coordconfskpos1pos2conf1conf2s r@rzAnnotator.kptss 88jj)..0DGZZ d"**GdODAq=Dt~~a'89'8!s1v'89&QR)G tQqTWGq!Q&7U1X+=+Bq6Q;Q4Dcz  477S\3w<$@&'SU`c`k`kl$ ::b>D"4==12D"Q%!)a013tRUQYN7K3LMD"Q%!)a013tRUQYN7K3LM19 "Q%!)a0E "Q%!)a0Es{eck 7U1X%*d1ga.@A.EaSTX\]^X_bcXc7U1X%*d1ga.@A.EaSTX\]^X_bcXc$tq?Q.R?Q!s1v?Q.R^_jmjujuv2 88 NN477 # 1:./Ss J4J9c@|jj||||y)z"Add rectangle to image (PIL-only).N)ryr)r=xyrrrs r@rzAnnotator.rectangles Bgu5rHc (|dk(r.|jj|\}}|dxxd|z z cc<|jr|rY|jj|\}}|jj |d|d|d|zdz|d|zdzf|d}d|vrm|j d}|jj|\} }|D]8} |jj || ||j|dxx|z cc<:y |jj ||||jy |rtj|d|j|jd\}}|d|z d k\} |d|z| r |d|z d z n |d|zd zf} tj|j|| |d tjd}tj|j||d|j||jtj y ) z'Adds text to an image using PIL or cv2.bottomr rrr2 rrrarrN)r]rroryrsplitrrrrrrwrr) r=rrranchor box_stylerrNlines_rrrs r@rzAnnotator.texts X 99$$T*DAq qEQUNE 88yy((.1 ##RUBqE2a519q="Q%!)a-$PW`#a+ t| 4(yy((.1!DIINN2t)$))NLqEQJE" r4idiiHtQ$''TWWUVWX1Q%!)q.UQY1 A beai!mK dggr2y"ckkJ+ KKr1dggyDGG^a^i^i jrHct|tjr|ntj||_t j |j|_y)z"Update self.im from a numpy array.N)rurrvrwrrxry)r=rws r@rvzAnnotator.fromarrays5"2u{{3"9LNN477+ rHc@tj|jS)z Return annotated image as array.)r9rrw)r=s r@resultzAnnotator.resultszz$''""rH)NNz Arial.ttfFabc))r+r+r+r2)rF))rrlT)NNr )r2topF) rRrSrTrUrArrrrrrvrrWrHr@rYrY<s4 g:2@"$H.$`6k:, #rHrYrWrc 4ddl}ddl}tjdtdtjdt t jd|dz d t|jd z}|dd }|j|gd  }|j|dddtdtdtj|dz dtjtj ddddd j#} | dj%|t'j(d||d zdz d} t+|D]=} | dj,| j/t1| Dcgc]}|dz  c}?| dj3d dt5|cxkrd!krXnnU| dj7t+t5|| dj9t;|j=d"d#$n| dj?d%|jAd&d'| ddd(|jA|d)d*| d+dd(d|ddddf<tCjD|d,z}tGjHt'jJd-t&jL.dz} tO|dd/|dd/D]5\}} tQjR| jU| d t1|07| d jW| | d jYd1d2D]*}d3D]#}| |jZ|j]d4%,|dz }tj|dtj|r ||yycc}w)5zCPlot training labels including class histograms and box statistics.rNignorez&The figure layout has changed to tight)categorymessage)rzPlotting labels to z labels.jpgz... r i@B)r_yrheight)columnsTautohist2binsg?)pmax)corner diag_kindkinddiag_kwsplot_kwszlabels_correlogram.jpgdpirC)rhrhfigsize tight_layoutr皙?)rrwidthr* instancesZrj)rotationfontsizeclassesr_r)r_raxrrrrra)r r rar3iroff)rr rCra)rrightleftrF)/pandasseabornwarningsfilterwarnings UserWarning FutureWarningrinforDrp DataFramepairplotdictpltsavefigclosesubplotsravelrr9linspacerangepatches set_colorr set_ylabelr7 set_xticksset_xticklabelslistr set_xlabelhistplotr xywh2xyxyrrvonesr;ziprrxrimshowaxisspines set_visible)boxesclsnamessave_diron_plotpdsnncr_r rrEimgrasfnames r@ plot_labelsr=s H{Dlm H}= KK%h&=%>dCD SWWY] B (7OE U$A BAKK$&vRT `djm`nKoKK33=IIK aF >q A G G IB 1 3R[[BQ7#=c JA 2Y ! Q!!F1I">Iq1s7I">?qE[!3u: 1s5z*+ 1d5<<>2R"M 1#KKSCBqEK=KKWRU#KFE!QqS&M MM% 4 'E //"''/BSH ICDS 5#;/Ss%%cF3K%H0qELLqEJJu 3A qELLO ' ' .4 | #EKK3IIK5#?s= N zim.jpggRQ?rjFTc>t|tjstj|}t j |j dd}|r5|ddddfjddjd|ddddf<|ddddf|z|z|ddddf<t j|j}t j||j|t|dt|dt|d t|d dd|rdndf} |ro|jjd d t!t#|j%d } t'j(| ddddfj+| dd| S)a Save image crop as {file} with crop size multiple {gain} and {pad} pixels. Save and/or return crop. This function takes a bounding box and an image, and then saves a cropped portion of the image according to the bounding box. Optionally, the crop can be squared, and the function allows for gain and padding adjustments to the bounding box. Args: xyxy (torch.Tensor or list): A tensor or list representing the bounding box in xyxy format. im (numpy.ndarray): The input image. file (Path, optional): The path where the cropped image will be saved. Defaults to 'im.jpg'. gain (float, optional): A multiplicative factor to increase the size of the bounding box. Defaults to 1.02. pad (int, optional): The number of pixels to add to the width and height of the bounding box. Defaults to 10. square (bool, optional): If True, the bounding box will be transformed into a square. Defaults to False. BGR (bool, optional): If True, the image will be saved in BGR format, otherwise in RGB. Defaults to False. save (bool, optional): If True, the cropped image will be saved to disk. Defaults to True. Returns: (numpy.ndarray): The cropped image. Example: ```python from ultralytics.utils.plotting import save_one_box xyxy = [50, 50, 150, 150] im = cv2.imread('image.jpg') cropped_im = save_one_box(xyxy, im, file='cropped.jpg', square=True) ``` rrPNrCr r)rr )rra)rr)rrCT)parentsexist_okz.jpg._)quality subsampling)rurrstackr xyxy2xywhviewrprr*long clip_boxesrsrDparentmkdirr|r with_suffixrrvsave) xyxyrwfilegainpadsquareBGRrLbcropfs r@ save_one_boxrV>sc> dELL ){{4  diiA&'A QU8<<?1%//2!QR%ABx$$AaeH ==  "DNN4" c$t*oc$t*o-s4:s4:/NPdX[STacPdd eD  $6 t$008 9 S$B$Y(--a-K KrHr3)rr-z images.jpgc )t|tjr,|jj j }t|tjr|jj }t|tjr|jj }t|tjr1|jj j t}t|tjr|jj }t|tjr|jj }d} d} |j\} } }}t| | } tj| dz)tj|ddkr|dz}tjt)|zt)|zdfdtj}t|D]U\}}|| k(rnKt||)zzt||)zz}}|j!dd d}|||||z|||zd d f<W| )z t||z }|dkrZt#j||z}t#j||z}t%j&|t))fd ||fD}t||z)zd z}t+|t-|d z |d|}t/dzD] }t||)zzt||)zz}}|j1||||z||zgd dd |r5|j3|dz|dzft5||j6d ddt9|dkDs||k(}||j d}t9|r-t;j<||d dfj>}|jddk(}|rd n||df}|jdr<|jdkr|dd gxx|zcc<|ddgxx|zcc<n |dkr||z}|dd gxx|z cc<|ddgxx|z cc<t|j>jAD]Z\}}||}tC|} |r|jE||n|}|s ||dkDs7|r|n |d||d}!|jG||!| \nKt9|r@|D];}tC|} |r|jE||n|}|j3||f|| d=t9|r||jI}"t9|"rQ|"djdks|"djdkr|"dxx|zcc<|"dxx|zcc<n |dkr|"|z}"|"dxx|z cc<|"dxx|z cc<t/t9|"D]!}s |dkDs|jK|"|#t9|s|jd|jdk(r||}#nr||g}#|jM}$tjN|$jQ|$ddfdz}%tjR|#|$d }#tjT|#|%k(d!d"}#tjV|jXjI}tj>jAD] \}}s |dkDstC||} |#|j\}&}'|&|k7s|'|k7rP|#|j tj}(t%j&|(||f}(|(j tZ}(n|#|j tZ}(t]j^t`5||||z|||zd d f|(d#ztjb| d$zz||||z|||zd d f|(<d d d |je| |jXjg|| r | |y y #1swYVxYw)%zPlot image grid with labels.irKrrr r*rar3rCNc3:K|]}t|zyw)NrL)rMr_nss r@rOzplot_images..s)Fv!#a"f+vsg{Gz?rjT)rrrorr2)rrl()r[r[)rrrDrPg)\(?g? z.1f)r)rr).r).r )r.g?gg?g333333?)4rurrrfloatrastyperDrsminr9ceilrpfullr;r transposemathrresizerQrYrqr!rrrnamer7r r*Trrgetrrrrrarangereshaperepeatwhererrwbool contextlibsuppressr}r:rvrL)*images batch_idxr2bboxesrrpathsr<r3r5max_size max_subplotsbsrrNrmosaicrErwr_rscalefs annotatoridxr r1labelsrjrr?rrkpts_ image_masksnlindexmhmwmaskrYs* @r@ plot_imagesrnsz&%,,'##%++-#u||$ggioo&%,,'##%%& !!#**3/$ %xxz!)U\\*MMO))+ HL,,KB1a R B s B vvfQiA# WWc"q&k3rAv;2Crxx HF6"2  1R=!3qAF|#41 \\!Q "&(qQw!a%"# # rMC1I %E qy IIeai  IIeai FE)F1v)F$FG a!er\D !B&U27^rt]bcI 1q5\1R=!3qAF|#41Q1q5!a%0$qQ  NNAE1q5>U1X0C0CCR0HTcN d s8a<q.C#hooe,G6{ fS"1"Wo688aA-%t6#q&>;;q>yy{d*q!f * q!f * q!f " q!f " '(89FAs A"1IE+0 !QaAa4*01#1T!WSM6J!++Ce+D :W A"1IE+0 !QaANNAq6aSUdNS! 4yS (u:V}((*d2eFm6G6G6IT6Qf * f * f " f " s5z*Aa4!uQx0+ 5z99Q<5;;q>1"'*K"'*KBIIbM112q!*=AE"$))K!"DK"$((;%+?c"JKZZ -224'(89FAsa4 &wqz 2!,Q!5!5B7bAg#.q>#8#8#BD#&::dQF#;D#';;t#4D#.q>#8#8#>D'00;<>qQw!a%QR?RD$(LL1LqLA1 AqD!A!%(1KK1%,,W51 1a166QST U1 1/;S\] ^1!r2 ) qELLN } $EKK3KIIK!2 D NN9!BqcB C C Ds+ H:)H5B(H:5H:: I)I$$I)c tj|||\}}}tt|D cgc]u} |t tj || |ddz |j ddz t tj || |ddz |j ddz fw} } tj||| |||ycc} w)a{ Plots a scatter plot with points colored based on a 2D histogram. Args: v (array-like): Values for the x-axis. f (array-like): Values for the y-axis. bins (int, optional): Number of bins for the histogram. Defaults to 20. cmap (str, optional): Colormap for the scatter plot. Defaults to 'viridis'. alpha (float, optional): Alpha for the scatter plot. Defaults to 0.8. edgecolors (str, optional): Edge colors for the scatter plot. Defaults to 'none'. Examples: >>> v = np.random.rand(100) >>> f = np.random.rand(100) >>> plt_color_scatter(v, f) rT)rr r)r?cmapr edgecolorsN) r9 histogram2dr!r7r_digitizersrscatter) vrUrrrrrxedgesyedgesrErs r@plt_color_scatterr1s&>>!QT:D&&Z__bcd_eYfhYfTU SQqT66:DJJqMAdtj@|j/d)}tj0|dtj2t5j6d|ycc}w)*a Plot the evolution results stored in an 'tune_results.csv' file. The function generates a scatter plot for each key in the CSV, color-coded based on fitness scores. The best-performing configurations are highlighted on the plots. Args: csv_file (str, optional): Path to the CSV file containing the tuning results. Defaults to 'tune_results.csv'. Examples: >>> plot_tune_results('path/to/tune_results.csv') rNrr r)rjrjTrviridisrnone)rrrzk+rd)rz = z.3grzri)fontdictbothrh)r. labelsizeztune_scatter_plots.pngrrzSaved )rjrfofitness)r linestylerrarrsmoothedrCrzFitness vs Iteration IterationFitnessztune_fitness.png)!rrrrrrrrr9argmaxrcr`r7rfigurersubplotrrrptitle tick_paramsyticks with_namerrrrr!xlabelylabelgridr)csv_filer6rrmnum_metrics_columnsr_keysrr|r8rErrmurNs r@plot_tune_resultsrMsS/H~H ;;x D#|| ,|!AGGI| ,-@-A BD A1gG 'A #d)s"#AJJxd3$1 a((( ) qT Aq!a% !W9B6R W[[]DR8 QCs2c(#vqk: Vq1 q5A: JJrN    6 7DKK#IIK KK&  aW!"AJJwT2HHQvYGHHQ!'3S VWXII$%JJ{JJyHHTNJJL   0 1DKK#IIK KK& I -sK;c g}t|D]\}}|d|ddfjjdd\}}}tj|j ddf|}|j tj||tj||fdtj|dj}|dddf|dddf|ddddffS)zZConvert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting.Nrf)rPr r r rrC) rrrrrarsappendcatr rEr) outputmax_dettargetsrErrrr2r|s r@output_to_targetrsG&!18G8RaR<,,.44YBT3 JJ 1 q)1 -uyy!S#--*? zrH)zpath/to/results.csvrFFFN)rrr)ztune_results.csv)r),rmrcrpathlibrrmatplotlib.pyplotpyplotrrr9rPILrrrrrultralytics.utilsrr r r r checksrrrrrrrrYr=rVzerosrr;rrrrrrrWrHr@rs=   ++*LL77!!E!EH G#G#T  "$tBx3 3l!%XTr%UZae-`  rxx4bhhq1RXXgRZZ8"@ @F;;|O87!t846EV@W@rH