- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I want to be able to set up ROI.
Documentation mention that it is possible when stacking up networks, I just want to do it for one network.
When I run this code:
InferenceEngine::ROI cropRoi; cropRoi.posX = inRoiBox.x; cropRoi.posY = inRoiBox.y; cropRoi.sizeX = inRoiBox.width; cropRoi.sizeY = inRoiBox.height; auto roiBlob = InferenceEngine::make_shared_blob(input, cropRoi); infer_request.SetBlob(item.first, roiBlob);
I get this exception: [Standard exception] Input blob size is not equal network input size (120000!=1555200).
Is it even possible to crop image to ROI if the network shape have been set during convertion?
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Michal,
Please refer to the object_detection_demo_ssd_async demo where the input image is scaled to the input model size.
Hope this helps.
Regards,
Ram prasad

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page