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
The Resize Image/Mask node provides multiple methods to change the dimensions of an input image or mask. It can scale by a multiplier, set specific dimensions, match the size of another input, or adjust based on pixel count, using various interpolation methods for quality.

Inputs

ParameterData TypeRequiredRangeDescription
inputIMAGE or MASKYesN/AThe image or mask to be resized.
resize_typeCOMBOYesSCALE_BY
SCALE_DIMENSIONS
SCALE_LONGER_DIMENSION
SCALE_SHORTER_DIMENSION
SCALE_WIDTH
SCALE_HEIGHT
SCALE_TOTAL_PIXELS
MATCH_SIZE
The method used to determine the new size. The required parameters change based on the selected type.
multiplierFLOATNo0.01 to 8.0The scaling factor. Required when resize_type is SCALE_BY (default: 1.00).
widthINTNo0 to 8192The target width in pixels. Required when resize_type is SCALE_DIMENSIONS or SCALE_WIDTH (default: 512).
heightINTNo0 to 8192The target height in pixels. Required when resize_type is SCALE_DIMENSIONS or SCALE_HEIGHT (default: 512).
cropCOMBONo"disabled"
"center"
The cropping method to apply when dimensions don’t match the aspect ratio. Only available when resize_type is SCALE_DIMENSIONS or MATCH_SIZE (default: “center”).
longer_sizeINTNo0 to 8192The target size for the longer side of the image. Required when resize_type is SCALE_LONGER_DIMENSION (default: 512).
shorter_sizeINTNo0 to 8192The target size for the shorter side of the image. Required when resize_type is SCALE_SHORTER_DIMENSION (default: 512).
megapixelsFLOATNo0.01 to 16.0The target total number of megapixels. Required when resize_type is SCALE_TOTAL_PIXELS (default: 1.0).
matchIMAGE or MASKNoN/AAn image or mask whose dimensions the input will be resized to match. Required when resize_type is MATCH_SIZE.
scale_methodCOMBOYes"nearest-exact"
"bilinear"
"area"
"bicubic"
"lanczos"
The interpolation algorithm used for scaling (default: “area”).
Note: The crop parameter is only available and relevant when the resize_type is set to SCALE_DIMENSIONS or MATCH_SIZE. When using SCALE_WIDTH or SCALE_HEIGHT, the other dimension is automatically scaled to maintain the original aspect ratio.

Outputs

Output NameData TypeDescription
resizedIMAGE or MASKThe resized image or mask, matching the data type of the input.