def NodeEvaluation(self, eval_info):
image1 = eval_info.EvaluateParameter('Image')
kernel_size = eval_info.EvaluateProperty('Kernel Size')
kernel_x = eval_info.EvaluateProperty('Kernel X')
kernel_y = eval_info.EvaluateProperty('Kernel Y')
filter_type = eval_info.EvaluateProperty('Filter Type')
image = api.RenderImage()
img = ArrayFromImage(image1.GetImage())
img = img.copy(order='C')
ctx = moderngl.create_standalone_context()
texture = ctx.texture(img.shape[1::-1], img.shape[2], img)
window_size = 3840 // 2, 2160 // 2
image_processing = ImageTransformer(ctx, window_size)
# Headless
image_processing.render(texture)
output_img = image_processing.write()
image.SetAsImage(ArrayToImage(output_img).convert('RGBA'))
self.NodeSetThumb(image.GetImage())
return image
def NodeEvaluation(self, eval_info):
image1 = eval_info.EvaluateParameter('Image')
sigma_val = eval_info.EvaluateProperty('Sigma')
intensity_val = eval_info.EvaluateProperty('Intensity')
# Convert the current image data to an array that scipy can use
im = ArrayFromImage(image1.GetImage())
# Create the image
if im.ndim == 3:
im_grey = np.zeros((im.shape[0], im.shape[1])).astype(float)
im_grey = (im[..., 0] * 0.3 + im[..., 1] * 0.6 + im[..., 2] * 0.1)
im = im_grey
im_smooth = self.SmoothGaussian(im, sigma_val)
sobel_x, sobel_y = self.Sobel(im_smooth)
# Calculate the normal map
generated_normal_map = self.ComputeNormalMap(
sobel_x,
sobel_y,
intensity_val
)
image = api.RenderImage()
image.SetAsImage(
ArrayToImage(generated_normal_map).convert('RGBA')
)
self.NodeSetThumb(image.GetImage())
return image
def NodeEvaluation(self, eval_info):
image1 = eval_info.EvaluateParameter('Image')
kernel_x = eval_info.EvaluateProperty('Kernel X')
kernel_y = eval_info.EvaluateProperty('Kernel Y')
filter_type = eval_info.EvaluateProperty('Filter Type')
image = api.RenderImage()
img = ArrayFromImage(image1.GetImage())
if filter_type == "Box":
output_img = cv2.boxFilter(img, -1, (kernel_x, kernel_y))
elif filter_type == "Gaussian":
# Both values must be odd
if (kernel_x % 2) == 0 and (kernel_y % 2) == 0:
kernel_y += 1
kernel_x += 1
output_img = cv2.GaussianBlur(img, (0, 0),
sigmaX=kernel_x,
sigmaY=kernel_y)
image.SetAsImage(ArrayToImage(output_img).convert('RGBA'))
self.NodeSetThumb(image.GetImage())
return image
def NodeEvaluation(self, eval_info):
image1 = eval_info.EvaluateParameter('Image')
saturation_val = eval_info.EvaluateProperty('Saturation')
brightness_val = eval_info.EvaluateProperty('Brightness')
gamma_val = eval_info.EvaluateProperty('Gamma')
# Convert the current image data to an array
# that we can use and greyscale it.
im = ArrayFromImage(image1.GetImage())
gray_scale_img = cv2.equalizeHist(cv2.cvtColor(im, cv2.COLOR_BGR2GRAY))
generated_roughness_map = self.ComputeRoughnessMap(
gray_scale_img, saturation_val, brightness_val, gamma_val)
image = api.RenderImage()
image.SetAsImage(ArrayToImage(generated_roughness_map).convert('RGBA'))
self.NodeSetThumb(image.GetImage())
return image
def NodeEvaluation(self, eval_info):
# TODO: Add a circle crop option
input_image = eval_info.EvaluateParameter('Image')
method = eval_info.EvaluateProperty("Method")
# Rectangle
x = eval_info.EvaluateProperty("X")
y = eval_info.EvaluateProperty("Y")
width = eval_info.EvaluateProperty("Width")
height = eval_info.EvaluateProperty("Height")
# Circle
center = eval_info.EvaluateProperty("Center")
radius = eval_info.EvaluateProperty("Radius")
input_image_array = ArrayFromImage(input_image.GetImage())
if method == "Rectangle":
output_image_array = input_image_array[y:y + height, x:x + width]
else:
# Create a blank mask
mask = np.zeros(shape=input_image_array.data.shape, dtype=np.uint8)
# Create the circle
circle_mask_colour = cv2.circle(mask, tuple(center), radius,
(255, 255, 255), -1)
# Convert the mask to gray scale
circle_mask = cv2.cvtColor(circle_mask_colour, cv2.COLOR_BGR2GRAY)
# Combine the mask and image
cropped_img = cv2.bitwise_and(input_image_array,
input_image_array,
mask=circle_mask)
# Crop the circle
circle_rect = cropped_img[center[1] - radius:center[1] + radius,
center[0] - radius:center[0] + radius]
output_image_array = circle_rect
image = api.RenderImage()
image.SetAsImage(ArrayToImage(output_image_array).convert("RGBA"))
self.NodeSetThumb(image.GetImage().convert("RGBA"))
return image
def NodeEvaluation(self, eval_info):
input_image = eval_info.EvaluateParameter("Image")
method = eval_info.EvaluateProperty("Method")
lower_threshold = eval_info.EvaluateProperty("Lower Threshold")
higher_threshold = eval_info.EvaluateProperty("Higher Threshold")
image = api.RenderImage()
# Consider removing the Pillow method?
if method == "Find Edges":
img = input_image.GetImage().convert("L").filter(
ImageFilter.FIND_EDGES)
image.SetAsImage(img.convert("RGBA"))
elif method == "Canny":
input_image_array = ArrayFromImage(input_image.GetImage())
output_image_array = cv2.Canny(input_image_array, lower_threshold,
higher_threshold)
image.SetAsImage(ArrayToImage(output_image_array).convert("RGBA"))
else:
image.SetAsImage(input_image.GetImage())
self.NodeSetThumb(image.GetImage())
return image
def NodeEvaluation(self, eval_info):
image1 = eval_info.EvaluateParameter('Image')
kernel_shape = eval_info.EvaluateProperty('Kernel Shape')
kernel_size = eval_info.EvaluateProperty('Kernel Size')
operation = eval_info.EvaluateProperty('Operation')
image = api.RenderImage()
img = ArrayFromImage(image1.GetImage())
if kernel_shape == "Rectangle":
kshape = cv2.MORPH_RECT
elif kernel_shape == "Ellipse":
kshape = cv2.MORPH_ELLIPSE
elif kernel_shape == "Cross":
kshape = cv2.MORPH_CROSS
kernel_img = cv2.getStructuringElement(kshape,
(kernel_size, kernel_size))
if operation == "Erode":
output_img = cv2.erode(img, kernel_img, iterations=1)
elif operation == "Dilate":
output_img = cv2.dilate(img, kernel_img, iterations=1)
elif operation == "Opening":
output_img = cv2.morphologyEx(img, cv2.MORPH_OPEN, kernel_img)
elif operation == "Closing":
output_img = cv2.morphologyEx(img, cv2.MORPH_CLOSE, kernel_img)
elif operation == "Top Hat":
output_img = cv2.morphologyEx(img, cv2.MORPH_TOPHAT, kernel_img)
elif operation == "Black Hat":
output_img = cv2.morphologyEx(img, cv2.MORPH_BLACKHAT, kernel_img)
image.SetAsImage(ArrayToImage(output_img).convert("RGBA"))
self.NodeSetThumb(image.GetImage())
return image