nodes/EllangoK-ComfyUI-post-processing-nodes
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

makes SineWave directional

Browse Source
This commit is contained in:
EllangoK
2023-05-08 20:39:54 -04:00
parent 01e0826bd6
commit 7dfa597cc3
2 changed files with 26 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
post_processing/sine_wave.py
+10 -4
View File
@@ -11,7 +11,7 @@ class SineWave:
"required": { "required": {
"image": ("IMAGE",), "image": ("IMAGE",),
"amplitude": ("FLOAT", { "amplitude": ("FLOAT", {
"default": 50, "default": 10,
"min": 0, "min": 0,
"max": 150, "max": 150,
"step": 5 "step": 5
@@ -22,6 +22,7 @@ class SineWave:
"max": 20, "max": 20,
"step": 1 "step": 1
}), }),
"direction": (["horizontal", "vertical"],),
}, },
} }
@@ -30,24 +31,29 @@ class SineWave:
CATEGORY = "postprocessing/Effects" CATEGORY = "postprocessing/Effects"
def apply_sine_wave(self, image: torch.Tensor, amplitude: float, frequency: float): def apply_sine_wave(self, image: torch.Tensor, amplitude: float, frequency: float, direction: str):
batch_size, height, width, channels = image.shape batch_size, height, width, channels = image.shape
result = torch.zeros_like(image) result = torch.zeros_like(image)
for b in range(batch_size): for b in range(batch_size):
tensor_image = image[b] tensor_image = image[b]
result[b] = self.sine_wave_effect(tensor_image, amplitude, frequency) result[b] = self.sine_wave_effect(tensor_image, amplitude, frequency, direction)
return (result,) return (result,)
def sine_wave_effect(self, image: torch.Tensor, amplitude: float, frequency: float): def sine_wave_effect(self, image: torch.Tensor, amplitude: float, frequency: float, direction: str):
height, width, _ = image.shape height, width, _ = image.shape
shifted_image = torch.zeros_like(image) shifted_image = torch.zeros_like(image)
for channel in range(3): for channel in range(3):
if direction == "horizontal":
for i in range(height): for i in range(height):
offset = int(amplitude * np.sin(2 * torch.pi * i * frequency / height)) offset = int(amplitude * np.sin(2 * torch.pi * i * frequency / height))
shifted_image[i, :, channel] = torch.roll(image[i, :, channel], offset) shifted_image[i, :, channel] = torch.roll(image[i, :, channel], offset)
elif direction == "vertical":
for j in range(width):
offset = int(amplitude * np.sin(2 * torch.pi * j * frequency / width))
shifted_image[:, j, channel] = torch.roll(image[:, j, channel], offset)
return shifted_image return shifted_image
post_processing_nodes.py
+10 -4
View File
@@ -1158,7 +1158,7 @@ class SineWave:
"required": { "required": {
"image": ("IMAGE",), "image": ("IMAGE",),
"amplitude": ("FLOAT", { "amplitude": ("FLOAT", {
"default": 50, "default": 10,
"min": 0, "min": 0,
"max": 150, "max": 150,
"step": 5 "step": 5
@@ -1169,6 +1169,7 @@ class SineWave:
"max": 20, "max": 20,
"step": 1 "step": 1
}), }),
"direction": (["horizontal", "vertical"],),
}, },
} }
@@ -1177,24 +1178,29 @@ class SineWave:
CATEGORY = "postprocessing/Effects" CATEGORY = "postprocessing/Effects"
def apply_sine_wave(self, image: torch.Tensor, amplitude: float, frequency: float): def apply_sine_wave(self, image: torch.Tensor, amplitude: float, frequency: float, direction: str):
batch_size, height, width, channels = image.shape batch_size, height, width, channels = image.shape
result = torch.zeros_like(image) result = torch.zeros_like(image)
for b in range(batch_size): for b in range(batch_size):
tensor_image = image[b] tensor_image = image[b]
result[b] = self.sine_wave_effect(tensor_image, amplitude, frequency) result[b] = self.sine_wave_effect(tensor_image, amplitude, frequency, direction)
return (result,) return (result,)
def sine_wave_effect(self, image: torch.Tensor, amplitude: float, frequency: float): def sine_wave_effect(self, image: torch.Tensor, amplitude: float, frequency: float, direction: str):
height, width, _ = image.shape height, width, _ = image.shape
shifted_image = torch.zeros_like(image) shifted_image = torch.zeros_like(image)
for channel in range(3): for channel in range(3):
if direction == "horizontal":
for i in range(height): for i in range(height):
offset = int(amplitude * np.sin(2 * torch.pi * i * frequency / height)) offset = int(amplitude * np.sin(2 * torch.pi * i * frequency / height))
shifted_image[i, :, channel] = torch.roll(image[i, :, channel], offset) shifted_image[i, :, channel] = torch.roll(image[i, :, channel], offset)
elif direction == "vertical":
for j in range(width):
offset = int(amplitude * np.sin(2 * torch.pi * j * frequency / width))
shifted_image[:, j, channel] = torch.roll(image[:, j, channel], offset)
return shifted_image return shifted_image