IssueTracker

Feature Request:

I would like to request an enhancement for the Google Cloud Vision API that allows users to specify a bounding box within an image and limit the detections and recognitions to the corresponding area within the image. This feature would be extremely useful for cases where the user is only interested in analyzing a specific part of an image, rather than the entire image.

What I would like to accomplish:

With this enhancement, I would be able to analyze images more efficiently and accurately by only focusing on the specific area of interest.

How this might work:

The enhanced API could include a new parameter that allows the user to specify the coordinates of a bounding box within the image. The API would then only analyze the specified region within the image, ignoring any detections or recognitions outside of that region.

If applicable, reasons why alternative solutions are not sufficient:

Currently, the only way to achieve this functionality is to load the entire image into memory, crop the image to the desired region, and then send the cropped image to the API for analysis. However, this workaround is not ideal as it requires additional processing and storage resources.

Other information:

The ability to specify a bounding box would be particularly useful when analyzing images stored in a Google Cloud Storage bucket, as it would allow users to analyze the images without having to download and crop them first.

Note:

It could be useful to consider the EXIF orientation flags that may be present in the image metadata too: https://issuetracker.google.com/issues/185418470