您好,登錄后才能下訂單哦!
本篇文章給大家分享的是有關如何在python中利用opencv實現一個膚色檢測功能,小編覺得挺實用的,因此分享給大家學習,希望大家閱讀完這篇文章后可以有所收獲,話不多說,跟著小編一起來看看吧。
原理:將RGB圖像轉換到YCRCB空間,膚色像素點會聚集到一個橢圓區域。先定義一個橢圓模型,然后將每個RGB像素點轉換到YCRCB空間比對是否再橢圓區域,是的話判斷為皮膚。
YCRCB顏色空間
橢圓模型
代碼
def ellipse_detect(image): """ :param image: 圖片路徑 :return: None """ img = cv2.imread(image,cv2.IMREAD_COLOR) skinCrCbHist = np.zeros((256,256), dtype= np.uint8 ) cv2.ellipse(skinCrCbHist ,(113,155),(23,15),43,0, 360, (255,255,255),-1) YCRCB = cv2.cvtColor(img,cv2.COLOR_BGR2YCR_CB) (y,cr,cb)= cv2.split(YCRCB) skin = np.zeros(cr.shape, dtype=np.uint8) (x,y)= cr.shape for i in range(0,x): for j in range(0,y): CR= YCRCB[i,j,1] CB= YCRCB[i,j,2] if skinCrCbHist [CR,CB]>0: skin[i,j]= 255 cv2.namedWindow(image, cv2.WINDOW_NORMAL) cv2.imshow(image, img) dst = cv2.bitwise_and(img,img,mask= skin) cv2.namedWindow("cutout", cv2.WINDOW_NORMAL) cv2.imshow("cutout",dst) cv2.waitKey()
效果
原理
針對YCRCB中CR分量的處理,將RGB轉換為YCRCB,對CR通道單獨進行otsu處理,otsu方法opencv里用threshold
代碼
def cr_otsu(image): """YCrCb顏色空間的Cr分量+Otsu閾值分割 :param image: 圖片路徑 :return: None """ img = cv2.imread(image, cv2.IMREAD_COLOR) ycrcb = cv2.cvtColor(img, cv2.COLOR_BGR2YCR_CB) (y, cr, cb) = cv2.split(ycrcb) cr1 = cv2.GaussianBlur(cr, (5, 5), 0) _, skin = cv2.threshold(cr1,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU) cv2.namedWindow("image raw", cv2.WINDOW_NORMAL) cv2.imshow("image raw", img) cv2.namedWindow("image CR", cv2.WINDOW_NORMAL) cv2.imshow("image CR", cr1) cv2.namedWindow("Skin Cr+OTSU", cv2.WINDOW_NORMAL) cv2.imshow("Skin Cr+OTSU", skin) dst = cv2.bitwise_and(img, img, mask=skin) cv2.namedWindow("seperate", cv2.WINDOW_NORMAL) cv2.imshow("seperate", dst) cv2.waitKey()
效果
原理
類似于第二種方法,只不過是對CR和CB兩個通道綜合考慮
代碼
def crcb_range_sceening(image): """ :param image: 圖片路徑 :return: None """ img = cv2.imread(image,cv2.IMREAD_COLOR) ycrcb=cv2.cvtColor(img,cv2.COLOR_BGR2YCR_CB) (y,cr,cb)= cv2.split(ycrcb) skin = np.zeros(cr.shape,dtype= np.uint8) (x,y)= cr.shape for i in range(0,x): for j in range(0,y): if (cr[i][j]>140)and(cr[i][j])<175 and (cr[i][j]>100) and (cb[i][j])<120: skin[i][j]= 255 else: skin[i][j] = 0 cv2.namedWindow(image,cv2.WINDOW_NORMAL) cv2.imshow(image,img) cv2.namedWindow(image+"skin2 cr+cb",cv2.WINDOW_NORMAL) cv2.imshow(image+"skin2 cr+cb",skin) dst = cv2.bitwise_and(img,img,mask=skin) cv2.namedWindow("cutout",cv2.WINDOW_NORMAL) cv2.imshow("cutout",dst) cv2.waitKey()
效果
原理
還是轉換空間然后每個通道設置一個閾值綜合考慮,進行二值化操作。
代碼
def hsv_detect(image): """ :param image: 圖片路徑 :return: None """ img = cv2.imread(image,cv2.IMREAD_COLOR) hsv=cv2.cvtColor(img,cv2.COLOR_BGR2HSV) (_h,_s,_v)= cv2.split(hsv) skin= np.zeros(_h.shape,dtype=np.uint8) (x,y)= _h.shape for i in range(0,x): for j in range(0,y): if(_h[i][j]>7) and (_h[i][j]<20) and (_s[i][j]>28) and (_s[i][j]<255) and (_v[i][j]>50 ) and (_v[i][j]<255): skin[i][j] = 255 else: skin[i][j] = 0 cv2.namedWindow(image, cv2.WINDOW_NORMAL) cv2.imshow(image, img) cv2.namedWindow(image + "hsv", cv2.WINDOW_NORMAL) cv2.imshow(image + "hsv", skin) dst = cv2.bitwise_and(img, img, mask=skin) cv2.namedWindow("cutout", cv2.WINDOW_NORMAL) cv2.imshow("cutout", dst) cv2.waitKey()
效果
示例
import cv2 import numpy as np def ellipse_detect(image): """ :param image: img path :return: None """ img = cv2.imread(image, cv2.IMREAD_COLOR) skinCrCbHist = np.zeros((256, 256), dtype=np.uint8) cv2.ellipse(skinCrCbHist, (113, 155), (23, 15), 43, 0, 360, (255, 255, 255), -1) YCRCB = cv2.cvtColor(img, cv2.COLOR_BGR2YCR_CB) (y, cr, cb) = cv2.split(YCRCB) skin = np.zeros(cr.shape, dtype=np.uint8) (x, y) = cr.shape for i in range(0, x): for j in range(0, y): CR = YCRCB[i, j, 1] CB = YCRCB[i, j, 2] if skinCrCbHist[CR, CB] > 0: skin[i, j] = 255 cv2.namedWindow(image, cv2.WINDOW_NORMAL) cv2.imshow(image, img) dst = cv2.bitwise_and(img, img, mask=skin) cv2.namedWindow("cutout", cv2.WINDOW_NORMAL) cv2.imshow("cutout", dst) cv2.waitKey() if __name__ == '__main__': ellipse_detect('./test.png')
以上就是如何在python中利用opencv實現一個膚色檢測功能,小編相信有部分知識點可能是我們日常工作會見到或用到的。希望你能通過這篇文章學到更多知識。更多詳情敬請關注億速云行業資訊頻道。
免責聲明:本站發布的內容(圖片、視頻和文字)以原創、轉載和分享為主,文章觀點不代表本網站立場,如果涉及侵權請聯系站長郵箱:is@yisu.com進行舉報,并提供相關證據,一經查實,將立刻刪除涉嫌侵權內容。