快捷導(dǎo)航

Python+Pillow+Pytesseract實現(xiàn)驗證碼識別

更新時間：2022年05月10日 15:00:37 作者：葉庭云

這篇文章主要為大家詳細介紹了如何利用pillow和pytesseract來實現(xiàn)驗證碼的識別，文中的示例代碼講解詳細，感興趣的小伙伴可以了解一下

一、環(huán)境配置

需要 pillow 和 pytesseract 這兩個庫，pip install 安裝就好了。

pip?install?pillow?-i?http://pypi.douban.com/simple?--trusted-host?pypi.douban.com
pip?install?pytesseract?-i?http://pypi.douban.com/simple?--trusted-host?pypi.douban.com

安裝好Tesseract-OCR.exe

pytesseract 庫的配置：搜索找到pytesseract.py，打開該.py文件，找到 tesseract_cmd，改變它的值為剛才安裝 tesseract.exe 的路徑。

二、驗證碼識別

識別驗證碼，需要先對圖像進行預(yù)處理，去除會影響識別準確度的線條或噪點，提高識別準確度。

實例1

import?cv2?as?cv
import?pytesseract
from?PIL?import?Image

def?recognize_text(image):
????#?邊緣保留濾波??去噪
????dst?=?cv.pyrMeanShiftFiltering(image,?sp=10,?sr=150)
????#?灰度圖像
????gray?=?cv.cvtColor(dst,?cv.COLOR_BGR2GRAY)
????#?二值化
????ret,?binary?=?cv.threshold(gray,?0,?255,?cv.THRESH_BINARY_INV?|?cv.THRESH_OTSU)
????#?形態(tài)學(xué)操作???腐蝕??膨脹
????erode?=?cv.erode(binary,?None,?iterations=2)
????dilate?=?cv.dilate(erode,?None,?iterations=1)
????cv.imshow('dilate',?dilate)
????#?邏輯運算??讓背景為白色??字體為黑??便于識別
????cv.bitwise_not(dilate,?dilate)
????cv.imshow('binary-image',?dilate)
????#?識別
????test_message?=?Image.fromarray(dilate)
????text?=?pytesseract.image_to_string(test_message)
????print(f'識別結(jié)果：{text}')


src?=?cv.imread(r'./test/044.png')
cv.imshow('input?image',?src)
recognize_text(src)
cv.waitKey(0)
cv.destroyAllWindows()

運行效果如下：

識別結(jié)果：3n3D
Process finished with exit code 0

實例2

import?cv2?as?cv
import?pytesseract
from?PIL?import?Image

def?recognize_text(image):
????#?邊緣保留濾波??去噪
????blur?=cv.pyrMeanShiftFiltering(image,?sp=8,?sr=60)
????cv.imshow('dst',?blur)
????#?灰度圖像
????gray?=?cv.cvtColor(blur,?cv.COLOR_BGR2GRAY)
????#?二值化
????ret,?binary?=?cv.threshold(gray,?0,?255,?cv.THRESH_BINARY_INV?|?cv.THRESH_OTSU)
????print(f'二值化自適應(yīng)閾值：{ret}')
????cv.imshow('binary',?binary)
????#?形態(tài)學(xué)操作??獲取結(jié)構(gòu)元素??開操作
????kernel?=?cv.getStructuringElement(cv.MORPH_RECT,?(3,?2))
????bin1?=?cv.morphologyEx(binary,?cv.MORPH_OPEN,?kernel)
????cv.imshow('bin1',?bin1)
????kernel?=?cv.getStructuringElement(cv.MORPH_OPEN,?(2,?3))
????bin2?=?cv.morphologyEx(bin1,?cv.MORPH_OPEN,?kernel)
????cv.imshow('bin2',?bin2)
????#?邏輯運算??讓背景為白色??字體為黑??便于識別
????cv.bitwise_not(bin2,?bin2)
????cv.imshow('binary-image',?bin2)
????#?識別
????test_message?=?Image.fromarray(bin2)
????text?=?pytesseract.image_to_string(test_message)
????print(f'識別結(jié)果：{text}')


src?=?cv.imread(r'./test/045.png')
cv.imshow('input?image',?src)
recognize_text(src)
cv.waitKey(0)
cv.destroyAllWindows()

運行效果如下：

二值化自適應(yīng)閾值：181.0
識別結(jié)果：8A62N1
Process finished with exit code 0

實例3

import?cv2?as?cv
import?pytesseract
from?PIL?import?Image

def?recognize_text(image):
????#?邊緣保留濾波??去噪
????blur?=?cv.pyrMeanShiftFiltering(image,?sp=8,?sr=60)
????cv.imshow('dst',?blur)
????#?灰度圖像
????gray?=?cv.cvtColor(blur,?cv.COLOR_BGR2GRAY)
????#?二值化??設(shè)置閾值??自適應(yīng)閾值的話?黃色的4會提取不出來
????ret,?binary?=?cv.threshold(gray,?185,?255,?cv.THRESH_BINARY_INV)
????print(f'二值化設(shè)置的閾值：{ret}')
????cv.imshow('binary',?binary)
????#?邏輯運算??讓背景為白色??字體為黑??便于識別
????cv.bitwise_not(binary,?binary)
????cv.imshow('bg_image',?binary)
????#?識別
????test_message?=?Image.fromarray(binary)
????text?=?pytesseract.image_to_string(test_message)
????print(f'識別結(jié)果：{text}')


src?=?cv.imread(r'./test/045.jpg')
cv.imshow('input?image',?src)
recognize_text(src)
cv.waitKey(0)
cv.destroyAllWindows()

運行效果如下：