![](/img/trans.png)
[英]Problem With Char Recognition pytesseract/tesseract-OCR/ Won't Recognize
[英]Pytesseract OCR doesn't recognize the digits
我正在嘗試閱讀這些圖片:
我嘗試了幾個選項,但我似乎無法將它們正確讀取為 15/0、30/0、40/0。
frame = frame[900:1000, 450:500]
scale_percent = 200 # percent of original size
width = int(frame.shape[1] * scale_percent / 100)
height = int(frame.shape[0] * scale_percent / 100)
dim = (width, height)
frame = cv2.resize(frame, dim, interpolation=cv2.INTER_AREA)
cv2.imshow("cropped", frame)
cv2.waitKey(0)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
cv2.imshow("cropped", frame)
cv2.waitKey(0)
pytesseract.pytesseract.tesseract_cmd = (
r"C:\Program Files\Tesseract-OCR\tesseract.exe"
)
results = pytesseract.image_to_data(
frame,
output_type=Output.DICT,
config="--psm 10 --oem 3 -c tessedit_char_whitelist=0123456789",
)
# results = replace_chars(results)
print(("-").join(results["text"]), "\n")
一種解決方法是使用inRange 閾值
結果將是:
如果你設置page-segmentation-mode 6
15
0
30
0
40
0
代碼:
import cv2
import pytesseract
from numpy import array
image_list = ["LZxCs.png", "W06I0.png", "vvzE5.png"]
for image in image_list:
bgr_image = cv2.imread(image)
hsv_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2HSV)
mask = cv2.inRange(hsv_image, array([0, 0, 0]), array([165, 10, 255]))
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 3))
dilate = cv2.dilate(mask, kernel, iterations=1)
thresh = cv2.bitwise_and(dilate, mask)
text = pytesseract.image_to_string(thresh, config='--psm 6')
print(text)
第二種方法是應用全局閾值
如果你設置page-segmentation-mode 6
15
0
30
0
40
0
代碼:
import cv2
import pytesseract
image_list = ["LZxCs.png", "W06I0.png", "vvzE5.png"]
for image in image_list:
bgr_image = cv2.imread(image)
gray_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray_image, 127, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
text = pytesseract.image_to_string(thresh, config='--psm 6')
print(text)
cv2.imwrite(f"/Users/ahx/Desktop/{image}", thresh)
cv2.imshow('', thresh)
cv2.waitKey(0)
有關更多信息,您可以查看文檔
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.