webcam_yolo_cuda.py

import numpy as np
import cv2
import os
import time
import signal
import timeit
import time
import sys
from termcolor import colored

cap = cv2.VideoCapture(0)

# Config
YOLO_DIRECTORY = "../models"
CONFIDENCE = 0.36
THRESHOLD = 0.22

ACTIVATION_RANGE = 800

# load the COCO class labels our YOLO model was trained on
labelsPath = os.path.sep.join([YOLO_DIRECTORY, "coco.names"])
LABELS = open(labelsPath).read().strip().split("\n")

# initialize a list of colors to represent each possible class label
np.random.seed(42)
COLORS = np.random.randint(0, 255, size=(len(LABELS), 3),
                           dtype="uint8")

# derive the paths to the YOLO weights and model configuration
weightsPath = os.path.sep.join([YOLO_DIRECTORY, "yolov3.weights"])
configPath = os.path.sep.join([YOLO_DIRECTORY, "yolov3.cfg"])

# Wait for buffering
time.sleep(0.4)

# load our YOLO object detector trained on COCO dataset (80 classes)
# and determine only the *output* layer names that we need from YOLO
print("[INFO] loading neural-network from disk...")
print(cv2.__version__)
net = cv2.dnn.readNetFromDarknet(configPath, weightsPath)

#net.setPreferableBackend(cv2.dnn.DNN_BACKEND_OPENCV)
#net.setPreferableTarget(cv2.dnn.DNN_TARGET_OPENCL)

net.setPreferableBackend(cv2.dnn.DNN_BACKEND_CUDA)
net.setPreferableTarget(cv2.dnn.DNN_TARGET_CUDA_FP16)

ln = net.getLayerNames()
ln = [ln[i[0] - 1] for i in net.getUnconnectedOutLayers()]
# loop over frames from the video file stream

# Handle Ctrl+C in terminal, release pointers
def signal_handler(sig, frame):
    # release the file pointers
    print("\n[INFO] cleaning up...")
    cv2.destroyAllWindows()
    sys.exit(0)

signal.signal(signal.SIGINT, signal_handler)

# Test for GPU support
build_info = str("".join(cv2.getBuildInformation().split()))
if cv2.ocl.haveOpenCL():
    cv2.ocl.setUseOpenCL(True)
    cv2.ocl.useOpenCL()
    print(colored("[OKAY] OpenCL is working!", "green"))
else:
    print(
        colored("[WARNING] OpenCL acceleration is disabled!", "yellow"))
if "CUDA:YES" in build_info:
    print(colored("[OKAY] CUDA is working!", "green"))
    #print(build_info)
else:
    print(
        colored("[WARNING] CUDA acceleration is disabled!", "yellow"))
print()

while(True):
    
    # Capture frame-by-frame
    start_time = timeit.default_timer()
    ret, frame = cap.read()
    
    H, W, channels = frame.shape
    
    # Our operations on the frame come here
    #gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    
    frame = cv2.cvtColor(frame, cv2.COLOR_RGBA2BGR)
    frame = cv2.UMat(frame)
    
    #32 64 96 128 160 192 224 256 288 320 352 384 416 448 480
    #512 544 576 608 640 672 704 736 768 800 832 864 896 928 960 992 1024
    blob = cv2.dnn.blobFromImage(frame, 1 / 260, (416, 416), swapRB=False, crop=False)
    net.setInput(blob)
    layerOutputs = net.forward(ln)
    
    boxes = []
    confidences = []
    classIDs = []
    
    # loop over each of the layer outputs
    for output in layerOutputs:
        # loop over each of the detections
        for detection in output:
            # extract the class ID and confidence (i.e., probability)
            # of the current object detection
            scores = detection[5:]
            classID = np.argmax(scores)
            # confidence = scores[classID]
            #classID = 2  # person = 0
            confidence = scores[classID]

            # filter out weak predictions by ensuring the detected
            # probability is greater than the minimum probability
            if confidence > CONFIDENCE:
                # scale the bounding box coordinates back relative to
                # the size of the image, keeping in mind that YOLO
                # actually returns the center (x, y)-coordinates of
                # the bounding box followed by the boxes' width and
                # height
                #box = detection[0: 4] = box.astype("int")
                
                
                box = detection[0: 4] * np.array([W, H, W, H])
                (centerX, centerY, width, height) = box.astype("int")

                x = int(centerX - (width / 2))
                y = int(centerY - (height / 2))
                

                boxes.append([x, y, int(width), int(height)])
                confidences.append(float(confidence))
                classIDs.append(classID)

    # apply non-maxima suppression to suppress weak, overlapping
    # bounding boxes
    idxs = cv2.dnn.NMSBoxes(boxes, confidences, CONFIDENCE, THRESHOLD)
    
    # ensure at least one detection exists
    if len(idxs) > 0:
        # Find best player match
        bestMatch = confidences[np.argmax(confidences)]

        # loop over the indexes we are keeping
        for i in idxs.flatten():
            # extract the bounding box coordinates
            (x, y) = (boxes[i][0], boxes[i][1])
            (w, h) = (boxes[i][2], boxes[i][3])

            # draw target dot on the frame
            #cv2.circle(frame, (int(x + w / 2), int(y + h / 5)), 5, (0, 0, 255), -1)

            # draw a bounding box rectangle and label on the frame
            color = [int(c) for c in COLORS[classIDs[i]]]
            cv2.rectangle(frame, (x, y),
                            (x + w, y + h), (0, 0, 255), 2)

            text = "TARGET {}%".format(int(confidences[i] * 100))
            cv2.putText(frame, text, (x, y - 5),
                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

    
    # Display the resulting frame
    cv2.imshow('frame',frame)
    
    elapsed = timeit.default_timer() - start_time
    sys.stdout.write(
        "\r{1} FPS with {0} MS interpolation delay \t".format(int(elapsed * 1000), int(1 / elapsed)))
    sys.stdout.flush()


    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

# When everything done, release the capture
cap.release()
cv2.destroyAllWindows()