Passing PIL cropped image as stream in Azure Computer Vision

Question

I need to pass a cropped image to the Computer Vision client in Azure, either as a stream or a url. I'm using PIL to crop the image based on some bounding box values. img1 stores the cropped image

img1 = img.crop((bbox[i][0], bbox[i][1], bbox[i][4], bbox[i][5]))
img1.show()
with open(img1.name, 'rb') as bg_image1:
     colour_analysis = computervision_client.analyze_image_in_stream(bg_image1, remote_image_features)

OR

img1 = img.crop((bbox[i][0], bbox[i][1], bbox[i][4], bbox[i][5]))
img1.show()
open_image = open(img1.name, "rb")
colour_analysis = computervision_client.analyze_image_in_stream(open_image, remote_image_features)

When I run the code above I get the error: raise AttributeError(name) AttributeError: name

Previously, when i uploaded the image from my files, I used this code which worked well:

with open('filepath\filename.jpg', 'rb') as bg_image:
    colour_analysis = computervision_client.analyze_image_in_stream(bg_image, remote_image_features)

How do I pass the cropped image to the Azure Client?

Answer 1

@Vedant Modi Thanks for the question. Can you please share the image that you are trying and SDK version that you are trying.

Please follow the below github sample to analyze a local image. See the ComputerVisionClient methods, such as AnalyzeImageInStreamAsync.

https://github.com/Azure-Samples/cognitive-services-quickstart-code/blob/master/dotnet/ComputerVision/ImageAnalysisQuickstart.cs#L306