Dasturiy injiniringi va raqamli iqtisodiyoti


Avtomatik iris holatini aniqlash


Download 65.06 Kb.
bet4/5
Sana11.11.2023
Hajmi65.06 Kb.
#1767144
1   2   3   4   5
Bog'liq
Ibrohimjon

Avtomatik iris holatini aniqlash. Algoritm chap va o'ng irislarning tasvirlarini ajratishga qodir.

  • Ishonchlilik. VeriEye algoritmi NIST IREX baholashlari davomida mukammal tanib olish aniqligini ko'rsatdi .

    Xulosa
    Ayni paytda iOS platformasida iris skanerini qo'llab-quvvatlash integratorlar tomonidan amalga oshirilishi kerak. Integratorlar shuni ta'kidlashlari kerakki, eng aniq ìrísí tanib olish faqat ìrísí tasvirlari yaqin infraqizil kameralar va tegishli yoritish bilan olingan taqdirdagina erishish mumkin. Biroq, irislarni smartfon yoki planshetlarda o'rnatilgan, to'g'ri yorug'lik va fokusdan foydalangan holda va to'g'ri muhitni tanlagan kameralar bilan suratga olinganda, hali ham oqilona aniqlik bilan tanib olish mumkin.
    Foydalanilgan adabiyotlar
    https://www.neurotechnology.com/verieye-system-requirements.html

    Amaliy qism:
    Code:
    #import the libraries
    import cv2 as cv
    import numpy as np
    #read the image
    img = cv.imread("D://medium_blogs//red_coat.jpg")
    #convert the BGR image to HSV colour space
    hsv = cv.cvtColor(img, cv.COLOR_BGR2HSV)
    #obtain the grayscale image of the original image
    gray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)
    #set the bounds for the red hue
    lower_red = np.array([160,100,50])
    upper_red = np.array([180,255,255])
    #create a mask using the bounds set
    mask = cv.inRange(hsv, lower_red, upper_red)
    #create an inverse of the mask
    mask_inv = cv.bitwise_not(mask)
    #Filter only the red colour from the original image using the mask(foreground)
    res = cv.bitwise_and(img, img, mask=mask)
    #Filter the regions containing colours other than red from the grayscale image(background)
    background = cv.bitwise_and(gray, gray, mask = mask_inv)
    #convert the one channelled grayscale background to a three channelled image
    background = np.stack((background,)*3, axis=-1)
    #add the foreground and the background
    added_img = cv.add(res, background)
    #create resizable windows for the images
    cv.namedWindow("res", cv.WINDOW_NORMAL)
    cv.namedWindow("hsv", cv.WINDOW_NORMAL)
    cv.namedWindow("mask", cv.WINDOW_NORMAL)
    cv.namedWindow("added", cv.WINDOW_NORMAL)
    cv.namedWindow("back", cv.WINDOW_NORMAL)
    cv.namedWindow("mask_inv", cv.WINDOW_NORMAL)
    cv.namedWindow("gray", cv.WINDOW_NORMAL)
    #display the images
    cv.imshow("back", background)
    cv.imshow("mask_inv", mask_inv)
    cv.imshow("added",added_img)
    cv.imshow("mask", mask)
    cv.imshow("gray", gray)
    cv.imshow("hsv", hsv)
    cv.imshow("res", res)
    if cv.waitKey(0):
    cv.destroyAllWindows()

    Download 65.06 Kb.

    Do'stlaringiz bilan baham:
  • 1   2   3   4   5




    Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
    ma'muriyatiga murojaat qiling