Skip to main content
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
Generate image using a control image (canny). This node takes a control image and generates a new image based on the provided prompt while following the edge structure detected in the control image.

Inputs

ParameterData TypeRequiredRangeDescription
control_imageIMAGEYes-The input image used for canny edge detection control
promptSTRINGNo-Prompt for the image generation (default: empty string)
prompt_upsamplingBOOLEANNo-Whether to perform upsampling on the prompt. If active, automatically modifies the prompt for more creative generation, but results are nondeterministic (same seed will not produce exactly the same result). (default: False)
canny_low_thresholdFLOATNo0.01 - 0.99Low threshold for Canny edge detection; ignored if skip_processing is True (default: 0.1)
canny_high_thresholdFLOATNo0.01 - 0.99High threshold for Canny edge detection; ignored if skip_processing is True (default: 0.4)
skip_preprocessingBOOLEANNo-Whether to skip preprocessing; set to True if control_image already is canny-fied, False if it is a raw image. (default: False)
guidanceFLOATNo1 - 100Guidance strength for the image generation process (default: 30)
stepsINTNo15 - 50Number of steps for the image generation process (default: 50)
seedINTNo0 - 18446744073709551615The random seed used for creating the noise. (default: 0)
Note: When skip_preprocessing is set to True, the canny_low_threshold and canny_high_threshold parameters are ignored since the control image is assumed to already be processed as a canny edge image.

Outputs

Output NameData TypeDescription
output_imageIMAGEThe generated image based on the control image and prompt