Single-File Processing Methods - CaptureVisionRouter Class

Method	Description
`capture`	Process an image or file to derive important information.
`capture_multi_pages`	Processes an image or file containing multiple pages to derive important information.

capture

Process an image or file to derive important information. It can optionally use a specified template for the capture.

def capture(self, *args) -> CapturedResult:

Parameters

args <tuple>: A variable-length argument list. Can be one of the following:

file_path <str>, template_name <str, optional>: Specifies the path of the file to process and a CaptureVisionTemplate to use for capturing.
file_bytes <bytes>, template_name <str, optional>: Specifies the image file bytes in memory to process and a CaptureVisionTemplate to use for capturing.
image_data <ImageData>, template_name <str, optional>: Specifies the image data to process and a CaptureVisionTemplate to use for capturing.
image <numpy.ndarray>, image_pixel_format <EnumImagePixelFormat, optional>, template_name <str, optional>: Specifies the image, formated in numpy.ndarray, and the image pixel format used in the image byte array to process and a CaptureVisionTemplate to use for capturing.

Remarks

There are two types of CaptureVisionTemplate: the preset ones which come with the SDK and the custom ones that get initialized when the user calls init_settings / init_settings_from_file.
When using a custom template, the parameter template_name should be the name of the CaptureVisionTemplate object in the JSON template file.
Please be aware that the preset CaptureVisionTemplates will be overwritten should the user call init_settings / init_settings_from_file and pass his own settings.
If parameter template_name is not specified, the preset one named ‘Default’ will be used. However, if the preset ones have been overwritten as described above, the first CaptureVisionTemplate from the user’s own settings will be used instead.

Return Value

Returns a CapturedResult object containing the captured items.

See Also

ImageData

Processes an image or file containing multiple pages to derive important information. It can optionally use a specified template for the capture.

def capture_multi_pages(self, file: Union[str, bytes], template_name: str = "") -> CapturedResultArray:

Parameters

file <Union[str, bytes]>: The input file containing multiple pages. It can be:
- A file path <str> pointing to a multi-page image (e.g., PDF, TIFF).
- A byte stream <bytes> representing the file content.
template_name <str, optional>: Specifies a CaptureVisionTemplate to use for capturing.

Remarks

There are two types of CaptureVisionTemplate: the preset ones which come with the SDK and the custom ones that get initialized when the user calls init_settings / init_settings_from_file.
When using a custom template, the parameter template_name should be the name of the CaptureVisionTemplate object in the JSON template file.
Please be aware that the preset CaptureVisionTemplates will be overwritten should the user call init_settings / init_settings_from_file and pass his own settings.
If parameter template_name is not specified, the preset one named ‘Default’ will be used. However, if the preset ones have been overwritten as described above, the first CaptureVisionTemplate from the user’s own settings will be used instead.

Return Value

Returns a CapturedResultArray object containing a collection of captured results, where each entry corresponds to a processed page.

See Also

CapturedResultArray