对于具有 table 结构的图像，使用 Pytesseract OCR 进行错误字符识别
Bad character recognition with Pytesseract OCR for images with table structure
我使用代码定位文本框并在它们周围创建一个矩形。这使我可以重建图像中 table 结构周围的网格。
然而，即使文本框检测工作得很好，如果我尝试定义每个矩形中存在的字符，pytesseract 不能很好地识别它们并且不允许找到原始文本。
这是我的 Python 代码：
import os
import cv2
import imutils
import argparse
import numpy as np
import pytesseract
# This only works if there's only one table on a page
# Important parameters:
# - morph_size
# - min_text_height_limit
# - max_text_height_limit
# - cell_threshold
# - min_columns
def pre_process_image(img, save_in_file, morph_size=(8, 8)):
# get rid of the color
pre = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
def img_estim(img, threshold=127):
is_dark = np.mean(img) < threshold
return True if is_dark else False
# Negative
if img_estim(pre):
print("non")
pre = cv2.bitwise_not(pre)
# Contrast & Brightness control
contrast = 2.0 #0 to 3
brightness = 0 #0 to 100
for y in range(pre.shape[0]):
for x in range(pre.shape[1]):
pre[y,x] = np.clip(contrast*pre[y,x] + brightness, 0, 255)
# Otsu threshold
pre = cv2.threshold(pre, 250, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
# dilate the text to make it solid spot
cpy = pre.copy()
struct = cv2.getStructuringElement(cv2.MORPH_RECT, morph_size)
cpy = cv2.dilate(~cpy, struct, anchor=(-1, -1), iterations=1)
pre = ~cpy
if save_in_file is not None:
cv2.imwrite(save_in_file, pre)
return pre
def find_text_boxes(pre, min_text_height_limit=15, max_text_height_limit=40):
# Looking for the text spots contours
# OpenCV 3
# img, contours, hierarchy = cv2.findContours(pre, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
# OpenCV 4
contours, hierarchy = cv2.findContours(pre, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
# Getting the texts bounding boxes based on the text size assumptions
boxes = []
for contour in contours:
box = cv2.boundingRect(contour)
h = box[3]
if min_text_height_limit < h < max_text_height_limit:
boxes.append(box)
return boxes
def find_table_in_boxes(boxes, cell_threshold=10, min_columns=2):
rows = {}
cols = {}
# Clustering the bounding boxes by their positions
for box in boxes:
(x, y, w, h) = box
col_key = x // cell_threshold
row_key = y // cell_threshold
cols[row_key] = [box] if col_key not in cols else cols[col_key] + [box]
rows[row_key] = [box] if row_key not in rows else rows[row_key] + [box]
# Filtering out the clusters having less than 2 cols
table_cells = list(filter(lambda r: len(r) >= min_columns, rows.values()))
# Sorting the row cells by x coord
table_cells = [list(sorted(tb)) for tb in table_cells]
# Sorting rows by the y coord
table_cells = list(sorted(table_cells, key=lambda r: r[0][1]))
return table_cells
def build_lines(table_cells):
if table_cells is None or len(table_cells) <= 0:
return [], []
max_last_col_width_row = max(table_cells, key=lambda b: b[-1][2])
max_x = max_last_col_width_row[-1][0] + max_last_col_width_row[-1][2]
max_last_row_height_box = max(table_cells[-1], key=lambda b: b[3])
max_y = max_last_row_height_box[1] + max_last_row_height_box[3]
hor_lines = []
ver_lines = []
for box in table_cells:
x = box[0][0]
y = box[0][1]
hor_lines.append((x, y, max_x, y))
for box in table_cells[0]:
x = box[0]
y = box[1]
ver_lines.append((x, y, x, max_y))
(x, y, w, h) = table_cells[0][-1]
ver_lines.append((max_x, y, max_x, max_y))
(x, y, w, h) = table_cells[0][0]
hor_lines.append((x, max_y, max_x, max_y))
return hor_lines, ver_lines
if __name__ == "__main__":
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", required=True,
help="path to input image to be OCR'd")
# ap.add_argument("-east", "--east", type=str,
# help="path to input EAST text detector")
args = vars(ap.parse_args())
in_file = os.path.join("images", args["image"])
pre_file = os.path.join("images", "pre.png")
out_file = os.path.join("images", "out.png")
img = cv2.imread(os.path.join(in_file))
top, bottom, left, right = [25]*4
img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_REPLICATE)
orig = img.copy()
pre_processed = pre_process_image(img, pre_file)
text_boxes = find_text_boxes(pre_processed)
cells = find_table_in_boxes(text_boxes)
hor_lines, ver_lines = build_lines(cells)
# (H, W) = img.shape[:2]
# net = cv2.dnn.readNet(args["east"])
# blob = cv2.dnn.blobFromImage(img, 1.0, (W, H),(123.68, 116.78, 103.94), swapRB=True, crop=False)
# net.setInput(blob)
# Visualize the result
vis = img.copy()
results = []
for box in text_boxes:
(x, y, w, h) = box
startX = x -2
startY = y -2
endX = x + w
endY = y + h
cv2.rectangle(vis, (startX, startY), (endX, endY), (0, 255, 0), 1)
roi=orig[startX:endX,startY:endY]
config = ("-l eng --psm 6")
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files (x86)\Tesseract-OCR\tesseract.exe'
text = pytesseract.image_to_string(roi,config=config )
results.append(((startX, startY, (endX), (endY)), text))
results = sorted(results, key=lambda r:r[0][1])
output = orig.copy()
for ((startX, startY, endX, endY), text) in results:
print("{}\n".format(text))
text = "".join([c if ord(c) < 128 else "" for c in text]).strip()
cv2.rectangle(output, (startX, startY), (endX, endY),(0, 0, 255), 1)
cv2.putText(output, text, (startX, startY - 20),cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 3)
# for line in hor_lines:
# [x1, y1, x2, y2] = line
# cv2.line(vis, (x1, y1), (x2, y2), (0, 0, 255), 1)
# for line in ver_lines:
# [x1, y1, x2, y2] = line
# cv2.line(vis, (x1, y1), (x2, y2), (0, 0, 255), 1)
cv2.imwrite(out_file, vis)
cv2.imshow("Text Detection", output)
cv2.waitKey(0)
初始图像：
Initial image
通过检测文本轮廓来定义矩形尺寸的预处理图像：
Preprocessed image with detection of text outlines to define the dimensions of rectangles
最终图像：
Final image
OCR 显示结果：
一个
拉
在
12
1
“
预先感谢您的帮助，希望我的描述足够清楚。
在执行 OCR 时，对图像进行预处理以获得 黑色前景文本和白色背景 是极其重要的。此外，放大图像有助于提高检测结果。我还发现在将其放入 Pytesseract 之前添加轻微的高斯模糊可以提高准确性。这是 --psm 6 将图像视为单个文本块的结果。查看 以获得更多配置选项。
经过预处理的放大、阈值处理和稍微模糊的图像
Pytesseract OCR 的结果
Series Type Scan Range CTDIvol DLP Phantom
(mm) (mGy) — (mGy-cm) cm
1 Scout - - - -
1 Scout - - - -
2 Axial = 113.554-1272.929 11.22 269.35 Body 32
Total Exam DLP: = 269.35
1/1
代码
import cv2
import pytesseract
import imutils
pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
image = cv2.imread('1.jpg')
image = imutils.resize(image, width=700)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
thresh = cv2.GaussianBlur(thresh, (3,3), 0)
data = pytesseract.image_to_string(thresh, lang='eng', config='--psm 6')
print(data)
cv2.imshow('thresh', thresh)
cv2.imwrite('thresh.png', thresh)
cv2.waitKey()
我使用代码定位文本框并在它们周围创建一个矩形。这使我可以重建图像中 table 结构周围的网格。
然而，即使文本框检测工作得很好，如果我尝试定义每个矩形中存在的字符，pytesseract 不能很好地识别它们并且不允许找到原始文本。
这是我的 Python 代码：
import os
import cv2
import imutils
import argparse
import numpy as np
import pytesseract
# This only works if there's only one table on a page
# Important parameters:
# - morph_size
# - min_text_height_limit
# - max_text_height_limit
# - cell_threshold
# - min_columns
def pre_process_image(img, save_in_file, morph_size=(8, 8)):
# get rid of the color
pre = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
def img_estim(img, threshold=127):
is_dark = np.mean(img) < threshold
return True if is_dark else False
# Negative
if img_estim(pre):
print("non")
pre = cv2.bitwise_not(pre)
# Contrast & Brightness control
contrast = 2.0 #0 to 3
brightness = 0 #0 to 100
for y in range(pre.shape[0]):
for x in range(pre.shape[1]):
pre[y,x] = np.clip(contrast*pre[y,x] + brightness, 0, 255)
# Otsu threshold
pre = cv2.threshold(pre, 250, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
# dilate the text to make it solid spot
cpy = pre.copy()
struct = cv2.getStructuringElement(cv2.MORPH_RECT, morph_size)
cpy = cv2.dilate(~cpy, struct, anchor=(-1, -1), iterations=1)
pre = ~cpy
if save_in_file is not None:
cv2.imwrite(save_in_file, pre)
return pre
def find_text_boxes(pre, min_text_height_limit=15, max_text_height_limit=40):
# Looking for the text spots contours
# OpenCV 3
# img, contours, hierarchy = cv2.findContours(pre, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
# OpenCV 4
contours, hierarchy = cv2.findContours(pre, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
# Getting the texts bounding boxes based on the text size assumptions
boxes = []
for contour in contours:
box = cv2.boundingRect(contour)
h = box[3]
if min_text_height_limit < h < max_text_height_limit:
boxes.append(box)
return boxes
def find_table_in_boxes(boxes, cell_threshold=10, min_columns=2):
rows = {}
cols = {}
# Clustering the bounding boxes by their positions
for box in boxes:
(x, y, w, h) = box
col_key = x // cell_threshold
row_key = y // cell_threshold
cols[row_key] = [box] if col_key not in cols else cols[col_key] + [box]
rows[row_key] = [box] if row_key not in rows else rows[row_key] + [box]
# Filtering out the clusters having less than 2 cols
table_cells = list(filter(lambda r: len(r) >= min_columns, rows.values()))
# Sorting the row cells by x coord
table_cells = [list(sorted(tb)) for tb in table_cells]
# Sorting rows by the y coord
table_cells = list(sorted(table_cells, key=lambda r: r[0][1]))
return table_cells
def build_lines(table_cells):
if table_cells is None or len(table_cells) <= 0:
return [], []
max_last_col_width_row = max(table_cells, key=lambda b: b[-1][2])
max_x = max_last_col_width_row[-1][0] + max_last_col_width_row[-1][2]
max_last_row_height_box = max(table_cells[-1], key=lambda b: b[3])
max_y = max_last_row_height_box[1] + max_last_row_height_box[3]
hor_lines = []
ver_lines = []
for box in table_cells:
x = box[0][0]
y = box[0][1]
hor_lines.append((x, y, max_x, y))
for box in table_cells[0]:
x = box[0]
y = box[1]
ver_lines.append((x, y, x, max_y))
(x, y, w, h) = table_cells[0][-1]
ver_lines.append((max_x, y, max_x, max_y))
(x, y, w, h) = table_cells[0][0]
hor_lines.append((x, max_y, max_x, max_y))
return hor_lines, ver_lines
if __name__ == "__main__":
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", required=True,
help="path to input image to be OCR'd")
# ap.add_argument("-east", "--east", type=str,
# help="path to input EAST text detector")
args = vars(ap.parse_args())
in_file = os.path.join("images", args["image"])
pre_file = os.path.join("images", "pre.png")
out_file = os.path.join("images", "out.png")
img = cv2.imread(os.path.join(in_file))
top, bottom, left, right = [25]*4
img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_REPLICATE)
orig = img.copy()
pre_processed = pre_process_image(img, pre_file)
text_boxes = find_text_boxes(pre_processed)
cells = find_table_in_boxes(text_boxes)
hor_lines, ver_lines = build_lines(cells)
# (H, W) = img.shape[:2]
# net = cv2.dnn.readNet(args["east"])
# blob = cv2.dnn.blobFromImage(img, 1.0, (W, H),(123.68, 116.78, 103.94), swapRB=True, crop=False)
# net.setInput(blob)
# Visualize the result
vis = img.copy()
results = []
for box in text_boxes:
(x, y, w, h) = box
startX = x -2
startY = y -2
endX = x + w
endY = y + h
cv2.rectangle(vis, (startX, startY), (endX, endY), (0, 255, 0), 1)
roi=orig[startX:endX,startY:endY]
config = ("-l eng --psm 6")
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files (x86)\Tesseract-OCR\tesseract.exe'
text = pytesseract.image_to_string(roi,config=config )
results.append(((startX, startY, (endX), (endY)), text))
results = sorted(results, key=lambda r:r[0][1])
output = orig.copy()
for ((startX, startY, endX, endY), text) in results:
print("{}\n".format(text))
text = "".join([c if ord(c) < 128 else "" for c in text]).strip()
cv2.rectangle(output, (startX, startY), (endX, endY),(0, 0, 255), 1)
cv2.putText(output, text, (startX, startY - 20),cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 3)
# for line in hor_lines:
# [x1, y1, x2, y2] = line
# cv2.line(vis, (x1, y1), (x2, y2), (0, 0, 255), 1)
# for line in ver_lines:
# [x1, y1, x2, y2] = line
# cv2.line(vis, (x1, y1), (x2, y2), (0, 0, 255), 1)
cv2.imwrite(out_file, vis)
cv2.imshow("Text Detection", output)
cv2.waitKey(0)
初始图像： Initial image 通过检测文本轮廓来定义矩形尺寸的预处理图像： Preprocessed image with detection of text outlines to define the dimensions of rectangles 最终图像： Final image OCR 显示结果：
一个 拉 在
12
1 “
预先感谢您的帮助，希望我的描述足够清楚。
在执行 OCR 时，对图像进行预处理以获得 黑色前景文本和白色背景 是极其重要的。此外，放大图像有助于提高检测结果。我还发现在将其放入 Pytesseract 之前添加轻微的高斯模糊可以提高准确性。这是 --psm 6 将图像视为单个文本块的结果。查看
经过预处理的放大、阈值处理和稍微模糊的图像
Pytesseract OCR 的结果
Series Type Scan Range CTDIvol DLP Phantom
(mm) (mGy) — (mGy-cm) cm
1 Scout - - - -
1 Scout - - - -
2 Axial = 113.554-1272.929 11.22 269.35 Body 32
Total Exam DLP: = 269.35
1/1
代码
import cv2
import pytesseract
import imutils
pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
image = cv2.imread('1.jpg')
image = imutils.resize(image, width=700)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
thresh = cv2.GaussianBlur(thresh, (3,3), 0)
data = pytesseract.image_to_string(thresh, lang='eng', config='--psm 6')
print(data)
cv2.imshow('thresh', thresh)
cv2.imwrite('thresh.png', thresh)
cv2.waitKey()