Anda harus menjalankan proses kamera setiap saat.
Ini adalah satu-satunya cara saya dapat mencapai hasil (rata-rata) 50 ms. Saya mencari solusi di mana-mana. 1 detik terlalu lambat untuk proyek sensor gerak saya.
Proyek @Dave Jones membantu saya mengetahui cara melakukannya.
Hanya 2 file:
sebuah daemon, berjalan sepanjang waktu dan klien.
Dasmon adalah tempat Anda mengatur semua pengaturan kamera.
picam-daemon.py
picam-client.py
python picam-daemon.py
import threading
import os, io, base64, time, socket, picamera, daemon
import daemon.runner
MAX_LENGTH = 50 # max length of any possible entry from "client"
serversocket = socket.socket(socket.AF_INET, socket.SOCK_STREAM) # setup socket
PORT = 10000 # port 10000
HOST = '127.0.0.1' # runs on local host
serversocket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1) # this allows us to override port, prevents error
serversocket.bind((HOST, PORT)) # lock server to this port and host
serversocket.listen(10) # max 10 clients
# Waits for commands, such as "snap" and "ack"
# Runs over "sockets"
def handle(clientsocket):
while 1:
buf = clientsocket.recv(MAX_LENGTH)
# Receive the SNAP command. Take a picture with PiCam.
if buf == 'snap':
start = time.time()
camera.capture('/home/pi/ir/picam-latest-snap.jpg')
finish = start - time.time()
print finish
print 'Picture Taken!'
if buf == 'ack':
print 'Ping: Hello!'
if len(buf) == 0: break
# Camera is always loaded here
# The "magic" is in the camThread, this allows a picture to be captured, then it gracefully closed the camera connection and reopens it. This produces very fast captures (54ms vs 1.5s!)
while 1:
# setup camera
camera = picamera.PiCamera()
camera.resolution = (640, 480)
#camera.zoom = (0.2, 0.2, 1.0, 1.0)
camera.exposure_mode = 'sports'
print('Camera server running')
# accept connections from outside, in order to receive commands
(clientsocket, address) = serversocket.accept()
ct = threading.Thread(target=handle, args=(clientsocket,))
ct.run() # this can be run(), because it can be scaled.
print 'Camera thread starting.'
camThread = threading.Thread()
while camThread.is_alive():
camThread.join(1)
camThread.run() # this must be start(), otherwise PiCam will crash. This is because PiCam cannot receive more than 1 connection.
print 'Camera thread ended'
camera.close() # Gracefully close PiCam if client disconnects
(di terminal kedua) python picam-client.py
import socket
import sys
HOST = '127.0.0.1'
PORT = 10000
s = socket.socket()
s.connect((HOST, PORT))
print s
while 1:
msg = raw_input("Command To Send: ")
if msg == "close":
s.close()
sys.exit(0)
s.send(msg)
Saya memposting jawaban ini karena saya menemukan ini di Google, mencoba menemukan jawaban sendiri. Saya tidak dapat menemukannya, jadi saya harus menggali beberapa proyek dan membuat sendiri sesuatu.