feat: Gaussian Blur Mask, Dilate Mask (SEGS), Gaussian Blur Mask (SEGS)
refactor: feather/gaussian blur fix: Mask To SEGS - invalid conversion when mask is donut shape fix: SEGS Paste/SEGS Preview - proper alpha handling
This commit is contained in:
Dr.Lt.Data
@@ -86,15 +86,20 @@ This custom node helps to conveniently enhance images through Detector, Detailer
|
||||
* SEGS Filter (range) - This node retrieves only SEGs from SEGS that have a size and position within a certain range.
|
||||
* SEGSConcat - Concatenate segs1 and segs2. If source shape of segs1 and segs2 are different from segs2 will be ignored.
|
||||
* Picker (SEGS) - Among the input SEGS, you can select a specific SEG through a dialog. If no SEG is selected, it outputs an empty SEGS. Increasing the batch_size of SEGSDetailer can be used for the purpose of selecting from the candidates.
|
||||
* Set Default Image For SEGS - Set a default image for SEGS. SEGS with images set this way do not need to have a fallback image set. When override is set to false, the original image is preserved.
|
||||
* DecomposeSEGS - Decompose SEGS to allow for detailed manipulation.
|
||||
* AssembleSEGS - Reassemble the decomposed SEGS.
|
||||
* From SEG_ELT - Extract detailed information from SEG_ELT.
|
||||
* Edit SEG_ELT - Modify some of the information in SEG_ELT.
|
||||
* Dilate SEG_ELT - Dilate the mask of SEG_ELT.
|
||||
* Set Default Image For SEGS - Set a default image for SEGS. SEGS with images set this way do not need to have a fallback image set. When override is set to false, the original image is preserved.
|
||||
* Dilate Mask (SEGS) - Dilate/Erosion Mask in SEGS
|
||||
* Gaussian Blur Mask (SEGS) - Apply Gaussian Blur to Mask in SEGS
|
||||
* SEGS_ELT Manipulation - experimental nodes
|
||||
* DecomposeSEGS - Decompose SEGS to allow for detailed manipulation.
|
||||
* AssembleSEGS - Reassemble the decomposed SEGS.
|
||||
* From SEG_ELT - Extract detailed information from SEG_ELT.
|
||||
* Edit SEG_ELT - Modify some of the information in SEG_ELT.
|
||||
* Dilate SEG_ELT - Dilate the mask of SEG_ELT.
|
||||
|
||||
* Dilate Mask - Dilate Mask.
|
||||
* Support erosion for negative value.
|
||||
* Mask Manipulation
|
||||
* Dilate Mask - Dilate Mask.
|
||||
* Support erosion for negative value.
|
||||
* Gaussian Blur Mask - Apply Gaussian Blur to Mask. You can utilize this for mask feathering.
|
||||
|
||||
* Pipe nodes
|
||||
* ToDetailerPipe, FromDetailerPipe - These nodes are used to bundle multiple inputs used in the detailer, such as models and vae, ..., into a single DETAILER_PIPE or extract the elements that are bundled in the DETAILER_PIPE.
|
||||
|
||||
@@ -213,6 +213,9 @@ NODE_CLASS_MAPPINGS = {
|
||||
"ImpactEdit_SEG_ELT": Edit_SEG_ELT,
|
||||
"ImpactDilate_Mask_SEG_ELT": Dilate_SEG_ELT,
|
||||
"ImpactDilateMask": DilateMask,
|
||||
"ImpactGaussianBlurMask": GaussianBlurMask,
|
||||
"ImpactDilateMaskInSEGS": DilateMaskInSEGS,
|
||||
"ImpactGaussianBlurMaskInSEGS": GaussianBlurMaskInSEGS,
|
||||
"ImpactScaleBy_BBOX_SEG_ELT": SEG_ELT_BBOX_ScaleBy,
|
||||
|
||||
"BboxDetectorCombined_v2": BboxDetectorCombined,
|
||||
@@ -375,6 +378,9 @@ NODE_DISPLAY_NAME_MAPPINGS = {
|
||||
"ImpactDilate_Mask_SEG_ELT": "Dilate Mask (SEG_ELT)",
|
||||
"ImpactScaleBy_BBOX_SEG_ELT": "ScaleBy BBOX (SEG_ELT)",
|
||||
"ImpactDilateMask": "Dilate Mask",
|
||||
"ImpactGaussianBlurMask": "Gaussian Blur Mask",
|
||||
"ImpactDilateMaskInSEGS": "Dilate Mask (SEGS)",
|
||||
"ImpactGaussianBlurMaskInSEGS": "Gaussian Blur Mask (SEGS)",
|
||||
|
||||
"PreviewBridge": "Preview Bridge (Image)",
|
||||
"PreviewBridgeLatent": "Preview Bridge (Latent)",
|
||||
|
||||
@@ -2,7 +2,7 @@ import configparser
|
||||
import os
|
||||
|
||||
|
||||
version = "V4.48.12"
|
||||
version = "V4.49"
|
||||
|
||||
dependency_version = 19
|
||||
|
||||
|
||||
@@ -971,8 +971,12 @@ def mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size=1, label='A',
|
||||
|
||||
else:
|
||||
mask_i_uint8 = (mask_i * 255.0).astype(np.uint8)
|
||||
contours, _ = cv2.findContours(mask_i_uint8, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
|
||||
for contour in contours:
|
||||
contours, ctree = cv2.findContours(mask_i_uint8, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
|
||||
for j, contour in enumerate(contours):
|
||||
hierarchy = ctree[0][j]
|
||||
if hierarchy[3] != -1:
|
||||
continue
|
||||
|
||||
separated_mask = np.zeros_like(mask_i_uint8)
|
||||
cv2.drawContours(separated_mask, [contour], 0, 255, -1)
|
||||
separated_mask = np.array(separated_mask / 255.0).astype(np.float32)
|
||||
|
||||
@@ -207,7 +207,7 @@ class DetailerForEach:
|
||||
else crop_ndarray4(image.numpy(), seg.crop_region)
|
||||
cropped_image = to_tensor(cropped_image)
|
||||
mask = to_tensor(seg.cropped_mask)
|
||||
mask = tensor_feather_mask(mask, feather)
|
||||
mask = tensor_gaussian_blur_mask(mask, feather)
|
||||
|
||||
is_mask_all_zeros = (seg.cropped_mask == 0).all().item()
|
||||
if is_mask_all_zeros:
|
||||
|
||||
@@ -238,7 +238,7 @@ class SEGSPaste:
|
||||
ref_tensor = ref_image_opt[i].unsqueeze(0)
|
||||
ref_image = crop_image(ref_tensor, seg.crop_region)
|
||||
if ref_image is not None:
|
||||
mask = tensor_feather_mask(seg.cropped_mask, feather, alpha/255)
|
||||
mask = tensor_gaussian_blur_mask(seg.cropped_mask, feather) * (alpha/255)
|
||||
x, y, *_ = seg.crop_region
|
||||
tensor_paste(image_i, ref_image, (x, y), mask)
|
||||
|
||||
@@ -305,7 +305,7 @@ class SEGSPreview:
|
||||
cropped_image = to_pil(cropped_image)
|
||||
|
||||
if alpha_mode:
|
||||
mask_array = seg.cropped_mask.astype(np.uint8) * 255
|
||||
mask_array = (seg.cropped_mask * 255).astype(np.uint8)
|
||||
mask_image = Image.fromarray(mask_array, mode='L').resize(cropped_image.size)
|
||||
cropped_image.putalpha(mask_image)
|
||||
|
||||
@@ -754,6 +754,78 @@ class DilateMask:
|
||||
return (torch.from_numpy(mask), )
|
||||
|
||||
|
||||
class GaussianBlurMask:
|
||||
@classmethod
|
||||
def INPUT_TYPES(s):
|
||||
return {"required": {
|
||||
"mask": ("MASK", ),
|
||||
"kernel_size": ("INT", {"default": 10, "min": 0, "max": 100, "step": 1}),
|
||||
"sigma": ("FLOAT", {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1}),
|
||||
}}
|
||||
|
||||
RETURN_TYPES = ("MASK", )
|
||||
|
||||
FUNCTION = "doit"
|
||||
|
||||
CATEGORY = "ImpactPack/Util"
|
||||
|
||||
def doit(self, mask, kernel_size, sigma):
|
||||
mask = torch.unsqueeze(mask, dim=-1)
|
||||
mask = utils.tensor_gaussian_blur_mask(mask, kernel_size, sigma)
|
||||
mask = torch.squeeze(mask, dim=-1)
|
||||
return (mask, )
|
||||
|
||||
|
||||
class DilateMaskInSEGS:
|
||||
@classmethod
|
||||
def INPUT_TYPES(s):
|
||||
return {"required": {
|
||||
"segs": ("SEGS", ),
|
||||
"dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}),
|
||||
}}
|
||||
|
||||
RETURN_TYPES = ("SEGS", )
|
||||
|
||||
FUNCTION = "doit"
|
||||
|
||||
CATEGORY = "ImpactPack/Util"
|
||||
|
||||
def doit(self, segs, dilation):
|
||||
new_segs = []
|
||||
for seg in segs[1]:
|
||||
mask = core.dilate_mask(seg.cropped_mask, dilation)
|
||||
seg = SEG(seg.cropped_image, mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper)
|
||||
new_segs.append(seg)
|
||||
|
||||
return ((segs[0], new_segs), )
|
||||
|
||||
|
||||
class GaussianBlurMaskInSEGS:
|
||||
@classmethod
|
||||
def INPUT_TYPES(s):
|
||||
return {"required": {
|
||||
"segs": ("SEGS", ),
|
||||
"kernel_size": ("INT", {"default": 10, "min": 0, "max": 100, "step": 1}),
|
||||
"sigma": ("FLOAT", {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1}),
|
||||
}}
|
||||
|
||||
RETURN_TYPES = ("SEGS", )
|
||||
|
||||
FUNCTION = "doit"
|
||||
|
||||
CATEGORY = "ImpactPack/Util"
|
||||
|
||||
def doit(self, segs, kernel_size, sigma):
|
||||
new_segs = []
|
||||
for seg in segs[1]:
|
||||
mask = utils.tensor_gaussian_blur_mask(seg.cropped_mask, kernel_size, sigma)
|
||||
mask = torch.squeeze(mask, dim=-1).squeeze(0).numpy()
|
||||
seg = SEG(seg.cropped_image, mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper)
|
||||
new_segs.append(seg)
|
||||
|
||||
return ((segs[0], new_segs), )
|
||||
|
||||
|
||||
class Dilate_SEG_ELT:
|
||||
@classmethod
|
||||
def INPUT_TYPES(s):
|
||||
|
||||
+11
-5
@@ -332,7 +332,7 @@ def _gaussian_kernel(kernel_size, sigma):
|
||||
return kernel / kernel.sum()
|
||||
|
||||
|
||||
def tensor_feather_mask(mask, thickness, base_alpha=1.0):
|
||||
def tensor_gaussian_blur_mask(mask, kernel_size, sigma=10.0):
|
||||
"""Return NHWC torch.Tenser from ndim == 2 or 4 `np.ndarray` or `torch.Tensor`"""
|
||||
if isinstance(mask, np.ndarray):
|
||||
mask = torch.from_numpy(mask)
|
||||
@@ -341,14 +341,20 @@ def tensor_feather_mask(mask, thickness, base_alpha=1.0):
|
||||
mask = mask[None, ..., None]
|
||||
_tensor_check_mask(mask)
|
||||
|
||||
if thickness <= 0:
|
||||
if kernel_size <= 0:
|
||||
return mask
|
||||
|
||||
# Create a feathered mask by applying a Gaussian blur to the mask
|
||||
mask = mask[:, None, ..., 0]
|
||||
prev_device = mask.device
|
||||
device = comfy.model_management.get_torch_device()
|
||||
mask.to(device)
|
||||
|
||||
blurred_mask = torchvision.transforms.GaussianBlur(kernel_size=thickness*2+1, sigma=10.0)(mask)
|
||||
# apply gaussian blur
|
||||
mask = mask[:, None, ..., 0]
|
||||
blurred_mask = torchvision.transforms.GaussianBlur(kernel_size=kernel_size*2+1, sigma=sigma)(mask)
|
||||
blurred_mask = blurred_mask[:, 0, ..., None]
|
||||
|
||||
blurred_mask.to(prev_device)
|
||||
|
||||
return blurred_mask
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user