Load Image Batch can't output mask

[akafun]

Load Image Batch can't output mask,Output mask node, useful in processing many png images.

At line 5071 of the source code

def load_batch_images(self, path, pattern='*', index=0, mode="single_image", label='Batch 001', allow_RGBA_output='false', filename_text_extension='true'):
        allow_RGBA_output = (allow_RGBA_output == 'true')
        if not os.path.exists(path):
            return (None, )
        fl = self.BatchImageLoader(path, label, pattern)
        new_paths = fl.image_paths
        if mode == 'single_image':
            image, filename = fl.get_image_by_id(index)
            if image == None:
                cstr(f"No valid image was found for the inded `{index}`").error.print()
                return (None, None)
        else:
            image, filename = fl.get_next_image()
            if image == None:
                cstr(f"No valid image was found for the next ID. Did you remove images from the source directory?").error.print()
                return (None, None)
        # Update history
        update_history_images(new_paths)
        if not allow_RGBA_output:
           image = image.convert("RGB")
        #addcodeSart
        if 'A' in image.getbands():
            mask = np.array(image.getchannel('A')).astype(np.float32) / 255.0
            mask = 1. - torch.from_numpy(mask)
        else:
            mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu")   
        #addcodeEnd   
        if filename_text_extension == "false":
            filename = os.path.splitext(filename)[0]
        return (pil2tensor(image),mask,filename)

Below is the relevant code, if I insert this code(#addcodeSart), will it become useful? Code from single image load Then return this mask and make sure RETURN_NAMES = ("image","mask","filename_text")

[WASasquatch]

That won't work if RGBA is not allowed, and should come before converting to RGB, that way an alpha channel can be scraped before it's truncated. Additionally masks are not batched aren't they? So maybe need a .unsqueeze(0)

[akafun]

That won't work if RGBA is not allowed, and should come before converting to RGB, that way an alpha channel can be scraped before it's truncated. Additionally masks are not batched aren't they? So maybe need a .unsqueeze(0)

Thanks, it worked