Live_Doc_Scanner.py

import cv2
import numpy as np

class docScanner :

    def __init__(self) -> None:
        pass

    def preprocessing(self,img):
        gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
        blur = cv2.GaussianBlur(gray,(7,7),5)
        canny = cv2.Canny(blur,100,0)
        kernel = np.ones((5,5))
        imgDial = cv2.dilate(canny,kernel,iterations=2)
        imgThre = cv2.erode(imgDial,kernel,iterations=1)
        return imgThre
    
    def getContours(self,img):
        biggest = np.array([])
        maxArea=0
        contours,heirarchy = cv2.findContours(img,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_NONE)
        for cnt in contours:
            area = cv2.contourArea(cnt)
            if area>3000:
                peri = cv2.arcLength(cnt,True)
                approx = cv2.approxPolyDP(cnt,0.02*peri,True)
                if area > maxArea and len(approx) == 4:
                    biggest = approx
                    maxArea = area
        cv2.drawContours(self.imgContour,biggest,-1,(255,0,0),20)
        return biggest

    def imgWarp(self,img,biggest):
        reordered = self.reorder(biggest)
        width,height = 480,640
        pts1 = np.float32(reordered)
        pts2 = np.float32([[0,0],[width,0],[0,height],[width,height]])
        matrix = cv2.getPerspectiveTransform(pts1,pts2)
        imgOut = cv2.warpPerspective(img,matrix,(width,height))
        return imgOut

    def reorder(self,mypoints):
        mypoints = mypoints.reshape((4,2))
        myPointsNew = np.zeros((4,1,2),np.int32)
        add = mypoints.sum(1)
        
        myPointsNew[0] = mypoints[np.argmin(add)]
        myPointsNew[3] = mypoints[np.argmax(add)]
        diff = np.diff(mypoints,axis=1)
        myPointsNew[1] = mypoints[np.argmin(diff)]
        myPointsNew[2] = mypoints[np.argmax(diff)]

        return myPointsNew

    def scan(self):
        cap = cv2.VideoCapture(0)
        while True:
            ret, frame = cap.read()
            imgThre = self.preprocessing(frame)
            self.imgContour = frame.copy()
            biggest = self.getContours(imgThre)
            if biggest.size ==0:
                cv2.imshow("Doc Scanner",self.imgContour)
            else :
                warped = self.imgWarp(frame,biggest)
                warped = warped[0:620,0:460]
                # a little sharpening the image if necessary
                # kernel = np.array([[0, -1, 0],
                #            [-1, 5,-1],
                #            [0, -1, 0]])
                # warped = cv2.filter2D(src=warped, ddepth=-1, kernel=kernel)
                cv2.imshow("DOC",warped)
                cv2.imshow("Doc Scanner",self.imgContour)
            if cv2.waitKey(20) & 0xFF == ord('q'):break    

        cap.release()
        cv2.destroyAllWindows()

Doc = docScanner()
Doc.scan()