mask crop square

mask_crop_square

功能1：输入图像和遮罩，按遮罩区域的中心裁剪图像和遮罩，输出为方形且固定大小的图像/遮罩/裁剪数据 注意1：若图像为多张，遮罩为1张则按这个遮罩裁剪所有图像 注意2：若图像为1张，遮罩为多张则按这些遮罩裁剪这个图像 注意3：若图像与遮罩批次均大于1且数量不一样则丢弃多的批次 注意4：若根据遮罩区域的中心计算的裁剪外框超出图像边界，则补充超出边界的部分为Exceeding指定的颜色(含图像和遮罩) 功能2：输入裁剪数据，将输入的图像按crop_data数据贴回原图像或图像批次 注意1：输入的任意图像都将拉伸缩放成原裁剪后的大小(无需再重缩放) 注意2：当输入crop_data时与其无关的参数都将无效 输入： images：输入图像或图像批次，若masks无输入则使用A通道裁剪 inversion_mask：是否翻转输入遮罩 mask_threshold：裁剪遮罩阈值，忽略浅色遮罩区域和零碎区域 size：指定输出大小(像素值)，若为0则使用输入遮罩区域的最大值 Exceeding: 超出裁剪范围的像素值填充方式, masks：输入用于获取裁剪数据的遮罩，若无输入则则使用images的A通道 若images输入无A通道则直接输出原图 crop_data：若此输入为空，则使用裁剪功能(功能1) 若此输入不为空，则将images贴回crop_data数据内含的原图上，此时有关于裁剪的参数将无效 option：用于输入裁剪设置，可忽略，目前仅增加防抖设置 输出： images：原输入图像或图像批次 masks：原输入遮罩或遮罩批次 crop_data：用于贴回原图像或图像批次，(包含图像大小/原图/位置信息) 若crop_data输入不为空则直接返回原crop_data输入 Function 1: Input images and masks, crop the images and masks based on the center of the mask region, and output square and fixed-sized images/masks/crop data. Note 1: If there are multiple images and only one mask, crop all images using this single mask. Note 2: If there is only one image but multiple masks, crop the image using these multiple masks. Note 3: If both the number of images and masks are greater than 1 but are not equal, discard the excess batches. Note 4: If the cropping box calculated from the center of the mask region exceeds the image boundary, fill the exceeded part with the color specified by "Exceeding" (for both image and mask). Function 2: Input crop data, and paste the input images back to the original image or batch of images based on the crop_data. Note 1: Any input image will be stretched and resized to the original cropped size (no need for further rescaling). Note 2: When inputting crop_data, all parameters unrelated to it will be invalid. Inputs: images: Input image or batch of images. If no masks are provided, use the A channel for cropping. inversion_mask: Whether to invert the input mask. mask_threshold: Threshold for cropping the mask, ignoring light-colored mask areas and fragmented regions. size: Specifies the output size (in pixels). If set to 0, use the maximum value of the input mask region. Exceeding: Method for filling pixels that exceed the cropping range. masks: Input masks used to obtain crop data. If no masks are provided, use the A channel of the images. If the images do not have an A channel, output the original image directly. crop_data: If this input is empty, use the cropping function (Function 1). If this input is not empty, paste the images back onto the original image contained in the crop_data. In this case, cropping parameters will be ignored. option: Used for input cropping settings, which can be ignored for now. Currently, only anti-shake settings are added. Outputs: images: The original input image or batch of images. masks: The original input mask or batch of masks. crop_data: Used for pasting back to the original image or batch of images (contains image size/original image/position information). If crop_data input is not empty, return the original crop_data input directly.

Pack: ComfyUI-WJNodes

custom_nodes.ComfyUI-WJNodes