run.py

from PIL import Image
import numpy as np
import cv2
from google.cloud import vision
import base64
from compare_faces import find

# os.system('export GOOGLE_APPLICATION_CREDENTIALS="/home/pi/Desktop/Assistant/gcp/lat.json"')

def findDistances(gun, people):
    print("===========================")
    (gXMin, gYMin, gXMax, gYMax) = gun

    gXCenter = (gXMax + gXMin / 2.)
    gYCenter = (gYMax + gYMin / 2.)
    print("gXCenter: ", gXCenter)
    print("gYCenter: ", gYCenter)

    distance = 10000000000.
    id = -1

    ROIxMin, ROIyMin, ROIxMax, ROIyMax = 0., 0., 0., 0.
    for (i, (pXMin, pYMin, pXMax, pYMax)) in enumerate(people):
        pXCenter = (pXMax + pXMin / 2.)
        pYCenter = (pYMax + pYMin / 2.)
        print("pXCenter: ", pXCenter)
        print("pYCenter: ", pYCenter)

        tmpX = (gXCenter - pXCenter) ** 2
        tmpY = (gYCenter - pYCenter) ** 2
        dist = (tmpX + tmpY) ** 0.5
        print("dist: ", dist)
        if dist < distance:
            distance = dist
            id = i
            ROIxMin, ROIyMin, ROIxMax, ROIyMax = pXMin, pYMin, pXMax, pYMax

    print("===========================")
    return id, distance, (ROIxMin, ROIyMin, ROIxMax, ROIyMax)


def localize_objects(path):
    """Localize objects in the local image.

    Args:
    path: The path to the local file.
    """
    client = vision.ImageAnnotatorClient()

    with open(path, 'rb') as image_file:
     content = image_file.read()
    # image = client.image(content=cv2.imencode('.jpg', path)[1].tostring())
    image = vision.types.Image(content=content)

    objects = client.object_localization(
        image=image).localized_object_annotations

    gun = tuple()
    people = []
    print('Number of objects found: {}'.format(len(objects)))
    for object_ in objects:
        print('\n{} (confidence: {})'.format(object_.name, object_.score))

        if object_.name == "Gun" or object_.name == "Weapon" or object_.name == "Handgun":
            print('Normalized bounding polygon vertices: ')
            xMin, ymin, xMax, yMax = 0.0, 0.0, 0.0, 0.0
            # tmpX = []
            # tmpY = []

            for (i, vertex) in enumerate(object_.bounding_poly.normalized_vertices):
                if i == 0:
                    xMin = vertex.x
                    yMin = vertex.y
                if i == 1:
                    xMax = vertex.x
                if i == 2:
                    yMax = vertex.y
                # tmpX.append(vertex.x)
                # tmpY.append(vertex.y)
                print(' - ({}, {})'.format(vertex.x, vertex.y))

            gun = (xMin, yMin, xMax, yMax)

        if object_.name == "Man" or object_.name == "Person" or object_.name == "Woman":
            print('Normalized bounding polygon vertices: ')
            xMin, ymin, xMax, yMax = 0.0, 0.0, 0.0, 0.0
            for (i, vertex) in enumerate(object_.bounding_poly.normalized_vertices):
                if i == 0:
                    xMin = vertex.x
                    yMin = vertex.y
                if i == 1:
                    xMax = vertex.x
                if i == 2:
                    yMax = vertex.y
                print(' - ({}, {})'.format(vertex.x, vertex.y))

            people.append((xMin, yMin, xMax, yMax))

    return gun, people



cap = cv2.VideoCapture(0)

while (1):
    print("in0")
    # Capture frame-by-frame
    ret, frame = cap.read()
    cv2.imshow("Live", frame)
    cv2.imwrite("test.jpg", frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

    print("in1")
    imagePath = "test.jpg"
    # img = cv2.imread("IMG_1944.jpeg")
    # print(type(img))
    # encoded = base64.b64encode(img)
    # print(encoded)

    # with open("IMG_1944.jpeg", 'rb') as image_file:
    #     content = image_file.read()
    #     print(content)
    # encoded = base64.b64encode(content)

    gun, people = localize_objects(imagePath)
    if not gun or not people:
        print("No threat detected")
        print(gun, people)
        pass
    else:
        print(gun)
        print(people)
        id, distance, (ROIxMin, ROIyMin, ROIxMax, ROIyMax)  = findDistances(gun, people)
        print(id, distance, ROIxMin, ROIyMin, ROIxMax, ROIyMax)

        im = Image.open(imagePath)
        width, height = im.size
        left = width * ROIxMin
        top = height - (height * ROIyMax)
        right = width * ROIxMax
        bottom = height - (height * ROIyMin)
        print(left, top, right, bottom)
        im1 = im.crop((left, top, right, bottom))
        # im1.show()

        im1.save('end.jpg')

        import pyimgur

        CLIENT_ID = "65865f9e351e0f1"
        PATH = "end.jpg"
        im = pyimgur.Imgur(CLIENT_ID)
        uploaded_image = im.upload_image(PATH, title="Uploaded with PyImgur")
        find("end.jpg")