How to extract the details of a bounding box detected in a RCNN trained network?

I am trying to detect arrow symbol from an image which is captured through a webcam. I am using RCNN to detect the arrows. I could detect the arrow from the image using a trained RCNN Network. The bounding box appears around the arrow.

I want to check using the details of bounding box, if a bounding box has been detected. My idea is to call a subfunction, if a bounding box has been detected in the image. How can i extract the details of bouding box and use it in my code for the above purpose.

Answer 1

[Raunak Gupta]

Hi,

After training the RCNN Detector you may getting a trainRCNNObjectDetector object which contains information about the detector. You may apply detect to the image for which object detection is required and get the bounding box information in terms of rectangle object. Also, the accuracy and labels are returned with the detect command.

You may refer to these examples.