Python+OpenCV實(shí)現(xiàn)閾值分割的方法詳解

第Python+OpenCV實(shí)現(xiàn)閾值分割的方法詳解目錄一、全局閾值1.效果圖2.源碼二、滑動(dòng)改變閾值(滑動(dòng)條)1.效果圖2.源碼三、自適應(yīng)閾值分割1.效果圖2.源碼3.GaussianBlur()函數(shù)去噪四、參數(shù)解釋

一、全局閾值

原圖：

整幅圖采用一個(gè)閾值，與圖片的每一個(gè)像素灰度進(jìn)行比較，重新賦值；

1.效果圖

2.源碼

importcv2

importmatplotlib.pyplotasplt

#設(shè)定閾值

thresh=130

#載入原圖，并轉(zhuǎn)化為灰度圖像

img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)

img_original=cv2.resize(img_original,(0,0),fx=0.3,fy=0.3)

#采用5種閾值類型（thresholdingtype）分割圖像

retval1,img_binary=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY)

retval2,img_binary_invertion=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY_INV)

retval3,img_trunc=cv2.threshold(img_original,thresh,255,cv2.THRESH_TRUNC)

retval4,img_tozero=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO)

retval5,img_tozero_inversion=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO_INV)

#采用plt.imshow()顯示圖像

imgs=[img_original,img_binary,img_binary_invertion,img_trunc,img_tozero,img_tozero_inversion]

titles=['original','binary','binary_inv','trunc','tozero','tozero_inv']

foriinrange(6):

plt.subplot(2,3,i+1)

plt.imshow(imgs[i],'gray')

plt.title(titles[i])

plt.xticks([])

plt.yticks([])

plt.show()

二、滑動(dòng)改變閾值(滑動(dòng)條)

1.效果圖

代碼如下（示例）：

importcv2

importnumpyasnp

importmatplotlib.pyplotasplt

#載入原圖，轉(zhuǎn)化為灰度圖像,并通過cv2.resize()等比調(diào)整圖像大小

img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)

img_original=cv2.resize(img_original,(0,0),fx=0.3,fy=0.3)

#初始化閾值，定義全局變量imgs

thresh=130

imgs=0

#創(chuàng)建滑動(dòng)條回調(diào)函數(shù)，參數(shù)thresh為滑動(dòng)條對應(yīng)位置的數(shù)值

defthreshold_segmentation(thresh):

#采用5種閾值類型（thresholdingtype）分割圖像

retval1,img_binary=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY)

retval2,img_binary_invertion=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY_INV)

retval3,img_trunc=cv2.threshold(img_original,thresh,255,cv2.THRESH_TRUNC)

retval4,img_tozero=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO)

retval5,img_tozero_inversion=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO_INV)

#由于cv2.imshow()顯示的是多維數(shù)組（ndarray）,因此我們通過np.hstack(數(shù)組水平拼接)

#和np.vstack(豎直拼接)拼接數(shù)組，達(dá)到同時(shí)顯示多幅圖的目的

img1=np.hstack([img_original,img_binary,img_binary_invertion])

img2=np.hstack([img_trunc,img_tozero,img_tozero_inversion])

globalimgs

imgs=np.vstack([img1,img2])

#新建窗口

dWindow('Images')

#新建滑動(dòng)條，初始位置為130

cv2.createTrackbar('thresholdvalue','Images',130,255,threshold_segmentation)

#第一次調(diào)用函數(shù)

threshold_segmentation(thresh)

#顯示圖像

while(1):

cv2.imshow('Images',imgs)

ifcv2.waitKey(1)==ord('q'):

break

cv2.destroyAllWindows()

三、自適應(yīng)閾值分割

1.效果圖

代碼如下（示例）：

importcv2

importmatplotlib.pyplotasplt

#載入原圖

img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)

#全局閾值分割

retval,img_global=cv2.threshold(img_original,130,255,cv2.THRESH_BINARY)

#自適應(yīng)閾值分割

img_ada_mean=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,15,3)

img_ada_gaussian=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)

imgs=[img_original,img_global,img_ada_mean,img_ada_gaussian]

titles=['OriginalImage','GlobalThresholding(130)','AdaptiveMean','AdaptiveGuassian',]

#顯示圖片

foriinrange(4):

plt.subplot(2,2,i+1)

plt.imshow(imgs[i],'gray')

plt.title(titles[i])

plt.xticks([])

plt.yticks([])

plt.show()

3.GaussianBlur()函數(shù)去噪

代碼如下（示例）：

importcv2

importmatplotlib.pyplotasplt

#載入原圖

img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)

#高斯濾波

img_blur=cv2.GaussianBlur(img_original,(13,13),13)#根據(jù)情況修改參數(shù)

#自適應(yīng)閾值分割

img_thresh=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)

img_thresh_blur=cv2.adaptiveThreshold(img_blur,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)

#顯示圖像

imgs=[img_thresh,img_thresh_blur]

titles=['img_thresh','img_thresh_blur']

foriinrange(2):

plt.subplot(1,2,i+1)

plt.imshow(imgs[i],'gray')

plt.title(titles[i])

plt.xticks([])

plt.yticks([])

plt.show()

四、參數(shù)解釋

1.cv2.threshold(src,thresh,maxval,type)

參數(shù)：

src:輸入的圖像

thresh:圖像分割所用的閾值（thresholdvalue）

maxval:當(dāng)閾值類型（thresholdingtype）采用cv2.THRESH_BINARY和cv2.THRESH_BINARY_INV時(shí)像素點(diǎn)被賦予的新值

type:介紹6種類型：

cv2.THRESH_BINARY（當(dāng)圖像某點(diǎn)像素值大于thresh(閾值)時(shí)賦予maxval，反之為0。注：最常用）

cv2.THRESH_BINARY_INV（當(dāng)圖像某點(diǎn)像素值小于thresh時(shí)賦予maxval，反之為0）

cv2.THRESH_TRUNC（當(dāng)圖像某點(diǎn)像素值大于thresh時(shí)賦予thresh,反之不變。注：雖然maxval沒用了，但是調(diào)用函數(shù)不能省略）

cv2.THRESH_TOZERO（當(dāng)圖像某點(diǎn)像素值小于thresh時(shí)賦予0，反之不變。注：同上）

cv2.THRESH_TOZ