Processing Takes Ridiculous time when OpenCV detects a positive image

user_16714154 0
Viewed 222

For some reason while I run OpenCV, when it detects a positive image instantly the detector processing time goes up ridiculously. Below in the code I labeled where the time is printed. Normally that output is about a tenth of a second but one it detects a positive image it instantly goes to 50 seconds and more! What causes this?

Note: This project is being done on a raspberry pi zero.

import picamera
from picamera.array import PiRGBArray
import cv2
import time
import numpy as np
import logging

image_path = '/home/pi/photo.bgr'


def DetectWeeds(img):
    time1 = time.time()
    weedDetector = cv2.CascadeClassifier('/home/pi/WeedClassifier/data/cascade.xml')

    found = weedDetector.detectMultiScale(img, minSize=(20, 20))
    time2 = time.time()
    print(time2 - time1)  #PROCESSING TIMER HERE

    amount_found = len(found)
    i = 0
    time.sleep(.1)
#Taking a lot of time when a weed is detected
    if amount_found != 0:
        # There may be more than one
        # sign in the image
        firstset = found[0]
        x, y, width, height = firstset

        imagewidth = 640

        currentx = (x + width) / 2 + x
        currenty = (y + height) / 2 + y

        targetx = imagewidth / 2

        return currentx, targetx, currenty

    if amount_found == 0:
        return 0, -1, 0

def main():
    camera = picamera.PiCamera()
    camera.resolution = (640, 480)
    camera.framerate = 16
    rawCapture = PiRGBArray(camera, size=(640, 480))
    for frame in camera.capture_continuous(rawCapture, format="bgr", use_video_port=True):

        image = frame.array

        currentx, targetx, currenty = DetectWeeds(image)

        if currentx != 0:

            if currentx > targetx:

                print("Move Left!")

            if currentx < targetx:
                print("Move Right!")

            if currentx == targetx:
                print("Centered!")
                print("Move Forward Now!")

        if currentx == 0 and targetx == -1:
            print("No Weeds Here")

        rawCapture.truncate(0)
        time.sleep(.08)
main()
raspberry-pi opencv python
2 Answers

I would suggest improving the code as below and see if you get any improvement in time. It can also be because when no weed is detected, the code is doing nothing and when weed is detected, the code has to run few more lines of code.

def DetectWeeds(img):
    time1 = time.time()
    weedDetector = cv2.CascadeClassifier('/home/pi/WeedClassifier/data/cascade.xml')

    found = weedDetector.detectMultiScale(img, minSize=(20, 20))
    time2 = time.time()
    print(time2 - time1)  #PROCESSING TIMER HERE

    amount_found = len(found)
    i = 0
    time.sleep(.1)
#Taking a lot of time when a weed is detected
    if amount_found != 0:
        # There may be more than one
        # sign in the image

        x, y, width, height = found[0]

        currentx = (x + width) / 2 + x
        currenty = (y + height) / 2 + y

        targetx = 640 / 2

        return currentx, targetx, currenty

    if amount_found == 0:
        return 0, -1, 0
user_102686210

found = weedDetector.detectMultiScale(img, minSize=(20, 20))

The problem is most likely lies in the scaleFactor parameter passed in detectMultiscale() function.

scaleFactor: Parameter specifying how much the image size is reduced at each image scale.

Tweak the scaleFactor in the detectMultiscale() function with multiple settings. Increasing the scale factor increases the accuracy and is also computationally expensive. The Good thumb rule is to find the right sweet spot where it runs in a reasonable time at sufficient accuracy.

If still if it doesn't work then try to adjust other parameters like maxSize,minNeighbors.

user_83364910
Related