mirror of
https://github.com/sml2h3/ddddocr.git
synced 2025-05-04 23:33:13 +08:00
52 lines
1.1 KiB
Python
52 lines
1.1 KiB
Python
import ddddocr
|
|
|
|
# # 识别
|
|
# ocr = ddddocr.DdddOcr(old=True)
|
|
#
|
|
# with open('test.jpg', 'rb') as f:
|
|
# img_bytes = f.read()
|
|
#
|
|
# res = ocr.classification(img_bytes)
|
|
# print(res)
|
|
|
|
# 目标检测
|
|
# det = ddddocr.DdddOcr(det=True)
|
|
#
|
|
# with open('微信图片_20220123122537.jpg', 'rb') as f:
|
|
# img_bytes = f.read()
|
|
#
|
|
# res = det.detection(img_bytes)
|
|
# print(res)
|
|
|
|
# 滑块模板匹配方式
|
|
import cv2
|
|
|
|
im = cv2.imread("b.jpg")
|
|
det = ddddocr.DdddOcr(det=False, ocr=False)
|
|
|
|
with open('a.jpg', 'rb') as f:
|
|
target_bytes = f.read()
|
|
|
|
with open('b.jpg', 'rb') as f:
|
|
background_bytes = f.read()
|
|
|
|
res = det.slide_match(target_bytes, background_bytes, simple_target=True)
|
|
print(res)
|
|
x = res["target"]
|
|
print(x)
|
|
im = cv2.line(im, (int(x[0]), 0), (int(x[0]), 300), color=(0, 0, 255), thickness=2)
|
|
cv2.imwrite("res.jpg", im)
|
|
|
|
# det = ddddocr.DdddOcr(det=False, ocr=False)
|
|
#
|
|
# with open('bg.jpg', 'rb') as f:
|
|
# target_bytes = f.read()
|
|
#
|
|
# with open('fullpage.jpg', 'rb') as f:
|
|
# background_bytes = f.read()
|
|
#
|
|
#
|
|
# res = det.slide_comparison(target_bytes, background_bytes)
|
|
#
|
|
# print(res)
|