/j .ddlmZddlZddlmZddlmZddlmZddl Z ddl Z d1d2dZ d3dZ d4dZd5dZd6dZd7d#Zd8d$Zd9d&Zd:d)Zd;d/Zd>> boxes = torch.tensor([[10, 10, 50, 50], [100, 100, 150, 150]]) >>> crop_box = [0, 0, 200, 200] >>> orig_box = [0, 0, 300, 300] >>> near_edge = is_box_near_crop_edge(boxes, crop_box, orig_box, atol=20.0) )dtypedeviceNr)r rtoldim)torch as_tensorrruncrop_boxes_xyxyisclose logical_andany)rr r r crop_box_torchorig_box_torchnear_crop_edgenear_image_edges _/home/longshao/multi-rider-rag/.venv/lib/python3.11/site-packages/ultralytics/models/sam/amg.pyis_box_near_crop_edger"s(_XU[VVVN_XU[VVVN eX . . 4 4 6 6E]5.qqq*ASTUUUNmE>$'+BTUVVVO&~7GHHN 9^ + + ++ batch_sizeintGenerator[list[Any]]c'(KrtfdDs Jdtdzttdzdkz}t|D]fdDVdS)aYield batches of data from input arguments with specified batch size for efficient processing. This function takes a batch size and any number of iterables, then yields batches of elements from those iterables. All input iterables must have the same length. Args: batch_size (int): Size of each batch to yield. *args (Any): Variable length input iterables to batch. All iterables must have the same length. Yields: (list[Any]): A list of batched elements from each input iterable. Examples: >>> data = [1, 2, 3, 4, 5] >>> labels = ["a", "b", "c", "d", "e"] >>> for batch in batch_iterator(2, data, labels): ... print(batch) [[1, 2], ['a', 'b']] [[3, 4], ['c', 'd']] [[5], ['e']] c3dK|]*}t|tdkV+dS)rN)len).0aargss r! z!batch_iterator..As8==1A#d1g,,.======r#z-Batched iteration must have same-size inputs.rc6g|]}|zdzzSr)r*argbr$s r! z"batch_iterator..Ds0JJJcs1z>QUj$889JJJr#N)allr)r%range)r$r, n_batchesr2s`` @r!batch_iteratorr7+s, nC=========nn?nnn =DG *ST!W 1Ja1O-P-PPI 9  KKJJJJJTJJJJJJJKKr#masksmask_thresholdthreshold_offsetc(|||zkdtjdtj}|||z kdtjdtj}||z S)aCompute the stability score for a batch of masks. The stability score is the IoU between binary masks obtained by thresholding the predicted mask logits at high and low values. Args: masks (torch.Tensor): Batch of predicted mask logits. mask_threshold (float): Threshold value for creating binary masks. threshold_offset (float): Offset applied to the threshold for creating high and low binary masks. Returns: (torch.Tensor): Stability scores for each mask in the batch. Examples: >>> masks = torch.rand(10, 256, 256) # Batch of 10 masks >>> mask_threshold = 0.5 >>> threshold_offset = 0.1 >>> stability_scores = calculate_stability_score(masks, mask_threshold, threshold_offset) Notes: - One mask is always contained inside the other. - Memory is saved by preventing unnecessary cast to torch.int64. )r)sumrint16int32)r8r9r: intersectionsunionss r!calculate_stability_scorerBGs0n/??@EEbPUP[E\\``ackpkv`wwM~(889 > >r > U U Y YZ\dido Y p pF 6 !!r# n_per_side np.ndarrayc dd|zz }tj|d|z |}tj|dddf|df}tj|dddfd|f}tj||gdddS)zaGenerate a 2D grid of evenly spaced points in the range [0,1]x[0,1] for image segmentation tasks.rNr<)axis)nplinspacetilestackreshape)rCoffsetpoints_one_sidepoints_xpoints_ys r!build_point_gridrQds !j. !Fk&!f*jAAOwtQQQw/*aAAHwqqq$w/!ZAAH 8Xx(r 2 2 2 : :2q A AAr#n_layersscale_per_layerlist[np.ndarray]cDfdt|dzDS)zPGenerate point grids for multiple crop layers with varying scales and densities.c Tg|]$}tt|zz %Sr0)rQr%)r*irCrSs r!r3z/build_all_layer_point_grids..os4 b b b S/A!BCC D D b b br#r)r5)rCrRrSs` `r!build_all_layer_point_gridsrXms0 b b b b beT\_`T`NaNa b b bbr#im_sizetuple[int, ...] overlap_ratio!tuple[list[list[int]], list[int]]c gg}}|\}}t||}|dd||g|dd}t|D]} d| dzz} t||zd| z z||| ||| fdt| D} fdt| D} t | | D]Z\} }| |t| z|t|z|g}|||| dz[||fS)aGenerate crop boxes of varying sizes for multiscale image processing, with layered overlapping regions. Args: im_size (tuple[int, ...]): Height and width of the input image. n_layers (int): Number of layers to generate crop boxes for. overlap_ratio (float): Ratio of overlap between adjacent crop boxes. Returns: crop_boxes (list[list[int]]): List of crop boxes in [x0, y0, x1, y1] format. layer_idxs (list[int]): List of layer indices corresponding to each crop box. Examples: >>> im_size = (800, 1200) # Height, width >>> n_layers = 3 >>> overlap_ratio = 0.25 >>> crop_boxes, layer_idxs = generate_crop_boxes(im_size, n_layers, overlap_ratio) rcBtj||dz z|z|z S)zZCalculate the length of each crop given the original length, number of crops, and overlap.r)mathceil)orig_lenn_cropsoverlaps r!crop_lenz%generate_crop_boxes..crop_lens%y'Wq[1H<GHHHr#rFrc:g|]}tz |zSr0r%)r*rWcrop_wrcs r!r3z'generate_crop_boxes..+TTTqsFW,122TTTr#c:g|]}tz |zSr0rf)r*rWcrop_hrcs r!r3z'generate_crop_boxes..rhr#)minappendr5r%r)rYrRr[ crop_boxes layer_idxsim_him_w short_siderdi_layern_crops_per_side crop_box_x0 crop_box_y0x0y0boxrjrgrcs @@@r!generate_crop_boxesryrs(  JJD$T4Jq!T4()))aIII??++1-mj0A8H4HIJJ$ 0'::$ 0'::TTTTTEBRr#pointsc|\}}}}tj||gg|j}t|jdkr|d}||zS)zAUncrop points by adding the crop box offset to their coordinates.r{r|rr})rr rvrwrrMs r! uncrop_pointsrs]LBAq \B8*V] ; ; ;F 6<A!!!$$ F?r#orig_horig_wc|\}}}}|dkr|dkr||kr||kr|S|||z z |||z z } }|||z || |z f} tjj|| dS)z]Uncrop masks by padding them to the original image size, handling coordinate transformations.r)value)rnn functionalpad) r8r rrrvrwx1y1pad_xpad_yrs r! uncrop_masksrsNBB Qww277rV||f  R"W%vb'95E urz2urz *C 8  " "5#Q " 7 77r#mask area_threshmodestrtuple[np.ndarray, bool]c ddl}|dvsJd|d|dk}||z tj}||d\}}}} |dddfd d} fd t | D} | s|d fSdg|  |s@ fd t |Dp$ttj| d zg tj | }|d fS)aRemove small disconnected regions or holes in a mask based on area threshold and mode. Args: mask (np.ndarray): Binary mask to process. area_thresh (float): Area threshold below which regions will be removed. mode (str): Processing mode, either 'holes' to fill small holes or 'islands' to remove small disconnected regions. Returns: processed_mask (np.ndarray): Processed binary mask with small regions removed. modified (bool): Whether any regions were modified. Examples: >>> mask = np.zeros((100, 100), dtype=np.bool_) >>> mask[40:60, 40:60] = True # Create a square >>> mask[45:55, 45:55] = False # Create a hole >>> processed_mask, modified = remove_small_regions(mask, 50, "holes") rN>holesislandszProvided mode z is invalidrr<rc,g|]\}}|k |dzSr/r0)r*rWsrs r!r3z(remove_small_regions..s&KKKtq!1{??QU???r#Fcg|]}|v| Sr0r0)r*rW fill_labelss r!r3z(remove_small_regions..s#JJJQQk5I5Iq5I5I5Ir#T) cv2astyperHuint8connectedComponentsWithStats enumerater5r%argmaxisin) rrrr correct_holes working_maskn_labelsregionsstatsrsizes small_regionsrs ` @r!remove_small_regionsrs7&JJJ ' ' ' ')K$)K)K)K ' ' 'GOM!D(00::L"%"B"Brss#""""" %%%%%% RV,,,,,:KKKK8"""":BBBBcccc 0"0"0"0"f8888""""J*E*E*E*E*E*Er#