Queue and Threading using Camera

Question

I have been using 2 threads in python and I have a global queue of size 1000.

1. For Grabbing the frames from camera - FrameGrabber Thread

   In this thread, I use OpenCV to grab the frame from camera and push it into the global queue. The camera works at the rate of 300 fps which means the thread capture frames for every 3.3 milli seconds.

   #Global queue
   frames = Queue.Queue(1000)
   
   class FrameGrabber(threading.Thread):
       def __init__(self, source):
            threading.Thread.__init__(self)
   
            self.ID = source
            self.cam = cv2.VideoCapture( source )
   
       def run(self):
            global frames
            ret, frame = self.cam.read()
            frames.put(frame)
   

2. For processing the frames grabbed - Processor Thread

   In this thread, I get the frames from the queue and do some process and this continues. The processing of one frame takes close to 4 milli seconds

   class Processor(threading.Thread):
       def __init__(self, source):
            threading.Thread.__init__(self)
   
       def run(self):
             global frames
             self.current_frame = frames.get()
             # Do some Processing
   

I am facing problems with this implementation

• I can see that some of the frames missing in the processor thread. We suspect that when the processor thread accessing the queue, the queue gets locked which is delaying the FrameGrabber thread to put frames into the queue.

Question: What is the data structure that can be the best fit for this scenario? I don't want to lose any frames from the camera ideally.