real time object detection?

[lezinterracial]

Using old haar cascades with opencv, but getting lots of false face detections. Got any other recommendations?

[pornmasta]

Do you use a threshold?
If yes be more strict

[SkynetHosting]

with a proper settings of tresholds you can get some better results.
but it is better to choose other methods if equipment afford that

[Colmike9]

If you aren't doing small objects, YOLO is a good object detection algorithm

[2MuchMark]

Try using Dlib’s face detector, MTCNN, or deep learning-based models such as OpenCV’s DNN module with pre-trained models (e.g., ResNet-based face detectors). Check scaleFactor and minNeighbors settings.

[lezinterracial]

Try using Dlib’s face detector, MTCNN, or deep learning-based models such as OpenCV’s DNN module with pre-trained models (e.g., ResNet-based face detectors). Check scaleFactor and minNeighbors settings.

Yea, I have read about Dlib. I want to use on my laptop. You think it will run with real time video.

Thanks for all responses so far.

[lezinterracial]

Try using Dlib’s face detector, MTCNN, or deep learning-based models such as OpenCV’s DNN module with pre-trained models (e.g., ResNet-based face detectors). Check scaleFactor and minNeighbors settings.

Yea, I have read about Dlib. I want to use on my laptop. You think it will run with real time video.

Thanks for all responses so far.

I used the example facedetect.py . Add a target box, when a face enters the box
moves a servo. Polulu Maestro servo controller.

import cv2
import os
import time

cmdpull = "/home/walter/maestro-linux/UscCmd --servo 5,7000"
cmdreturn = "/home/walter/maestro-linux/UscCmd --servo 5,6000"

# Load the cascade
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')

# To capture video from webcam. 
cap = cv2.VideoCapture(2)
# To use a video file as input 
# cap = cv2.VideoCapture('filename.mp4')

while True:
    # Read the frame
    _, img = cap.read()
    # Convert to grayscale
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # Detect the faces
    faces = face_cascade.detectMultiScale(gray, 1.1, 4)
   #create region of interest box and dot
    cv2.rectangle(img, (100, 100), (180, 180), (0, 0, 255), 2)
    cv2.line(img, (140, 140), (140, 140), (0, 255, 0), 5) 
    # Draw the rectangle around each face
    for (x, y, w, h) in faces:
        cv2.rectangle(img, (x, y), (x+w, y+h), (255, 0, 0), 2)
    #if face is in middle of the screen - region of interest
        if x > 100 and x < 180 and y > 100 and y < 180:
            print ("x: ",x)
            print ("y: ",y)
            time.sleep(.1)
            os.system(cmdpull)
            time.sleep(.1)
            os.system(cmdreturn)
    # Display
    cv2.imshow('img', img)
    # Stop if escape key is pressed
    k = cv2.waitKey(30) & 0xff
    if k==27:
        break
# Release the VideoCapture object
cap.release()

[lezinterracial]

If you aren't doing small objects, YOLO is a good object detection algorithm

Looking at YOLOv8 tutorials now. Thanks.

Scratch that for now. Will need a gpu. Having fun watching tutorials though. Now I know what Elon and that other guy were talking about when they mention Distributed Neural Network.