使用Python每5秒在每个项目上并行运行一个函数

Question

I am processing a video and want to check multiple regions of interest in parallel, every 5 seconds. I would like to know if the function I currently have is accomplishing this in the correct manner.

def checkFrames():
    timer = threading.Timer(5.0, checkFrames)
    timer.daemon = True
    timer.start()
    if(started):
        index = 0
        for region in regions:
            threading.Thread(target=processRegion, args=(frame, region, index)).start()
            index += 1

Right now I can't really tell if all of the regions are being processed in parallel (although I know that they are all being processed). The program runs pretty slow here compared to when I'm purposely checking the regions right after the other. The processing for all the regions don't have to start at the same time.

Answer 1

是的，此代码似乎每5秒并行启动N个线程，但是由于Python的全局解释器锁 ，它的执行速度不会比检查N个区域的单个线程快。

Answer 2

You are currently creating 6 new threads for each iteration (1 checkFrames and 5 processRegion ). Creating a threads is time and resources consuming. What I suggest is to create 5 threads that will process each every 5 seconds

def checkFrame(frame, region, index):
    while True:
        start_time = time.time()
        processRegion(frame, region ,index)
        sleep(5 - time.time() - start_time)
for region in len(regions):
    threading.Thread(target=checkFrame, args=(frame, region, index)).start()

This code assumes that a processRegion call takes under 5 seconds.