How to Extract Text from Scanned Books with Python and OpenCV

### Import OpenCV for image processing and NumPy for array operations. import cv2 import numpy as np  ### Read the input document image from disk. ### Provide a proper path to a scanned page or book photo. img = cv2.imread('Open-CV/Words-Segmentation/book2.jpg')  ### Define a function that prepares a binary mask where text is white on black. ### Steps: convert to grayscale, invert-binary threshold to flip text to white, smooth with Gaussian blur, and re-threshold to clean edges. def thresholding(image):     ### Convert the BGR image to grayscale to simplify intensity operations.     img_gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)      ### Apply inverse binary threshold so darker text becomes white.     ### Values below 170 turn to 255 (white), others to 0 (black).     # We need the words in white      ret, thresh = cv2.threshold(img_gray, 170, 255 , cv2.THRESH_BINARY_INV)          ### Smooth the mask to merge small gaps between characters for stronger word blobs.     # wee need the "merge" the characters to a single "words"     thresh = cv2.GaussianBlur(thresh, (11,11), 0)      ### Re-threshold the blurred mask to get a crisp binary image again.     ret, thresh = cv2.threshold(thresh, 130 , 255 , cv2.THRESH_BINARY)      ### Return the clean binary mask where words are white.     return thresh  ### Create the thresholded mask of the input image for downstream segmentation. thresh_img = thresholding(img)  ### Visualize the binary mask to verify the text is white and background is black. cv2.imshow("thresh_img",thresh_img)  ### Prepare containers and a rectangular kernel to detect whole text lines. # line sementation linesArray = [] kernelRows = np.ones((5,40), np.uint8)  ### Dilate horizontally to connect characters and words into long line components. # We will use dilation for the line sementation dilated = cv2.dilate(thresh_img, kernelRows, iterations=1)  ### Preview the dilated image to see connected line regions. cv2.imshow("dilated",dilated)  ### Extract external contours which should correspond to line blobs after dilation. # find contours (contoursRows , heirarchy) = cv2.findContours(dilated.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)  ### Iterate over detected contours, filter by area, and store bounding boxes for candidate lines. #cv2.drawContours(img, contoursRows, -1 , (0,255,0), 2) # draw the contour around each row on the original image # loop inside the contours and draw rectangle for row in contoursRows :     ### Compute contour area to filter out small noise.     area = cv2.contourArea(row)      ### Keep only sufficiently large components which likely represent text lines.     if area > 500 :         ### Compute the bounding rectangle of the contour.         x , y, w, h = cv2.boundingRect(row)          ### Optionally draw the line rectangle on the original image for inspection.         #cv2.rectangle(img, (x,y), (x+w, y+h), (40,100,250), 2)          ### Save line bounding box for sorted processing.         linesArray.append([x , y, w, h])  ### Print how many line candidates were found as a quick sanity check. print(len(linesArray)) # 33 lines  ### Sort lines from top to bottom using the y coordinate. # lets sort the line by the y position (from up to down) sortedLinesArray = sorted(linesArray, key=lambda line : line [1])  ### Prepare structures for word extraction and indexing. # words segmentation words = [] lineNumber = 0 all = []  ### Create a square kernel to connect characters into words without bridging adjacent lines. # kernel for words kernelWords = np.ones((7,7), np.uint8)  ### Dilate the thresholded image with the word kernel to merge letters into word-sized blobs. dilateWordsImg = cv2.dilate(thresh_img, kernelWords, iterations=1)  ### Show the dilation result used for word segmentation. cv2.imshow("dilate Words Img",dilateWordsImg)  ### For each detected line, focus on its region and find word contours inside it. for line in sortedLinesArray :      ### Unpack line rectangle.     x,y,w,h = line      ### Crop the region of interest for the current line to avoid cross-line interference.     roi_line = dilateWordsImg[y: y+h , x:x+w]      ### Optionally draw the line on the original image for debugging.     #cv2.rectangle(img, (x,y), (x+w, y+h), (0,255,0),2 )      ### Find external contours within the line ROI which correspond to merged words.     (contoursWords , heirarchy) = cv2.findContours(roi_line.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)      ### For each word contour, compute its bounding box and draw it on the original image.     for word in contoursWords:         x1,y1,w1,h1 = cv2.boundingRect(word)          ### Draw the word rectangle in yellow to visualize segmentation quality.         cv2.rectangle(img,(x+x1,y+y1),(x+x1+w1,y+y1+h1),(255,255,0),2)           ### Save absolute coordinates for later sorting and indexing.         words.append([x+x1,y+y1,x+x1+w1,y+y1+h1])      ### Sort all accumulated words by the left x coordinate so reading order is preserved.     # sort the words by the X position     sortedWords = sorted(words, key=lambda line : line[0]) #(x,y,w,h)      ### Build a list of (line_index, word_bbox) pairs representing a reading-order structure.     # build a full array of lines and words     for word in sortedWords:         a = (lineNumber,word)         all.append(a)      ### Increment line index for the next iteration.     lineNumber = lineNumber + 1  ### Optionally print the structure for debugging or downstream processing. #print(all)  ### Select an example word box from the combined list and print its coordinates. # show the first word in the first row chooseWord = all[3][1] print(chooseWord)  ### Extract the chosen word as a ROI from the original image to verify segmentation. roiWord = img[chooseWord[1]:chooseWord[3], chooseWord[0]:chooseWord[2]] cv2.imshow("Show a word", roiWord)  ### Show the original image with all drawn word rectangles for final inspection. cv2.imshow("Show the words",img)  ### Save the visualized segmentation to disk for documentation or OCR handoff. cv2.imwrite("c:/temp/segmentedBook.png",img)  ### Wait for a key press to keep windows open, then close all GUI windows cleanly. cv2.waitKey(0) cv2.destroyAllWindows()