cv_script.py

import cv2 as cv
import numpy as np
# from matplotlib import pyplot as plt


def image_resize(image, width = None, height = None, inter = cv.INTER_AREA):
    # initialize the dimensions of the image to be resized and
    # grab the image size
    dim = None
    (h, w) = image.shape[:2]

    # if both the width and height are None, then return the
    # original image
    if width is None and height is None:
        return image

    # check to see if the width is None
    if width is None:
        # calculate the ratio of the height and construct the
        # dimensions
        r = height / float(h)
        dim = (int(w * r), height)

    # otherwise, the height is None
    else:
        # calculate the ratio of the width and construct the
        # dimensions
        r = width / float(w)
        dim = (width, int(h * r))

    # resize the image
    resized = cv.resize(image, dim, interpolation = inter)

    # return the resized image
    return resized


# def resize_v2():

def produce_output(target_name):  
    target = cv.imread(r'./temp/target.jpg', 1)
    target = image_resize(target, height=500, width=650)
    # target = cv.resize(target, (650, 650))
    target2 = np.zeros((1080, 1728, 3), np.uint8)
    print(target2.shape)
    print(target.shape)
    offset_y = round((target2.shape[0]/2 - target.shape[0]/2))
    offset_x = round((target2.shape[1]/2 - target.shape[1]/2)) 
    print(offset_y, offset_x)
    target2[160:160+target.shape[0], 505: 505+target.shape[1]] = target
    # cv.imshow('target', target2)


    src = cv.imread(r'./resources/base.jpg')
    imgRGB = cv.cvtColor(src, cv.COLOR_BGR2RGB)
    imgHSV = cv.cvtColor(imgRGB, cv.COLOR_BGR2HSV)

    lower = np.array([60, 240, 245])    #Lower values of HSV range; Green have Hue value equal 120, but in opencv Hue range is smaler [0-180]
    upper = np.array([70, 255, 255])  #Uppervalues of HSV range
    imgRange = cv.inRange(imgHSV, lower, upper)

    # kernels for morphology operations
    kernel_noise = np.ones((3,3),np.uint8) #to delete small noises
    kernel_dilate = np.ones((9,9),np.uint8)  #bigger kernel to fill holes after ropes
    kernel_erode = np.ones((1,1),np.uint8)  #bigger kernel to delete pixels on edge that was add after dilate function

    imgErode = cv.erode(imgRange, kernel_noise, 1)
    imgDilate = cv.dilate(imgErode , kernel_dilate, 1)
    imgErode = cv.erode(imgDilate, kernel_erode, 1)


    mask = cv.cvtColor(imgErode, cv.COLOR_GRAY2BGR)
    res = cv.subtract(src, mask)
    # # cv.imshow('res', res)

    # mask of target
    res_t = cv.bitwise_and(target2, target2, mask = imgErode)

    res_f = cv.add(res, res_t)
    # cv.imshow('final', res_f)
    cv.imwrite('./temp/output.jpg', res_f)
    # # cv.imshow('res_t', res_t)