def GetText(path=None,result_pic=False):
"""
文字识别
:param path:图片路径
:return:
"""
if not path:
path = "F:/git/wxbot/dm/pic/222.png"
# ocr = PaddleOCR(use_gpu=False,det_model_dir="./inference/ch_det_mv3_db/",rec_model_dir="./inference/ch_det_mv3_db/") # need to run only once to download and load model into memory
ocr = PaddleOCR(gpu=False, det_model_dir="./inference/ch_det_r50_vd_db",
rec_model_dir="./inference/ch_rec_r34_vd_crnn_infer") # need to run only once to download and load model into memory
result = ocr.ocr(path)
txts = [line[1][0] for line in result]
# 展示图片
if result_pic:
image = Image.open(path).convert('RGB')
boxes = [line[0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores)
im_show = Image.fromarray(im_show)
TMP_FILE = os.path.join(conf.TMP_PATH, "result.png")
im_show.save(TMP_FILE)
# print(txts)
return txts
async def tt_ocr(request):
try:
# 获取数据
req_json = request.json
imgbase64 = req_json['imageBase64']
do_img = base64_to_image(imgbase64)
# 处理图片
img_result = ocr.ocr(do_img)
# 画图
boxes = [line[0] for line in img_result]
txts = [line[1][0] for line in img_result]
scores = [line[1][1] for line in img_result]
done_img = draw_ocr(
do_img,
boxes,
txts,
scores,
font_path='/path/to/PaddleOCR/doc/fonts/simfang.ttf')
# out图片转base64
imgbase64 = image_to_base64(done_img)
ret_data = {'out_imageBase64': imgbase64}
return response.json(ret_data)
except:
traceback.print_exc()
def ocr_result_image(result, input_image, output_image='output_image.png'):
if not os.path.exists('./simfang.ttf'):
os.system(f"wget -q {fds_url}/data/simfang.ttf")
image = Image.open(input_image).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path="./simfang.ttf")
# im_show = draw_ocr(image, boxes, txts, scores, font_path=get_module_path("../data/simfang.ttf", __file__))
im_show = Image.fromarray(im_show)
im_show.save(output_image)
return output_image
def drawImg(result, image):
image = safeShape(image)
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image,
boxes,
txts,
scores,
font_path="/code/doc/simfang.ttf")
im_show = Image.fromarray(im_show)
im_show.save(base_path + '/static/result.jpg')
return toPngStr(im_show)
def ocr_result_image(result, input_image, output_image='output_image.png'):
# https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/develop/doc/simfang.ttf
image = Image.open(input_image).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image,
boxes,
txts,
scores,
font_path=get_module_path("../data/simfang.ttf",
__file__))
im_show = Image.fromarray(im_show)
im_show.save(output_image)
return output_image
def imageOcrOperator():
image_path = "./static/images/input.jpg"
out_path = './static/images/orc_result.jpg'
result = OCR.ocr(image_path)
image = Image.open(image_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image,
boxes,
txts,
scores,
font_path='./static/tools/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save(out_path)
return out_path
def detect_OCR(self, img):
# Paddleocr目前支持中英文、英文、法语、德语、韩语、日语，可以通过修改lang参数进行切换
# 参数依次为`ch`, `en`, `french`, `german`, `korean`, `japan`。
ocr = PaddleOCR(use_angle_cls=True, use_gpu=False, lang="ch", use_space_char=True,det_db_unclip_ratio=2.0) # det_db_unclip_ratio=2.5这是参数定义位置，这个参数是检测后处理时控制文本框大小的，默认2.0，可以尝试改成2.5或者更大，反之，如果觉得文本框不够紧凑，也可以把该参数调小。 need to run only once to download and load model into memory
img_path = img
result = ocr.ocr(img_path,cls=False)
for line in result:
print(line)
# 显示结果
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='/path/to/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
return boxes, txts
def ocr_once():
# Paddleocr目前支持中英文、英文、法语、德语、韩语、日语，可以通过修改lang参数进行切换
# 参数依次为`ch`, `en`, `french`, `german`, `korean`, `japan`。
ocr = PaddleOCR(
use_angle_cls=True, lang="ch"
) # need to run only once to download and load model into memory
img_path = 'test/3.jpg'
img = cv2.imread(img_path)
start = time.time()
result = ocr.ocr(img, cls=True)
print(result)
print(time.time() - start)
# 显示结果
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result/result3.jpg')
import os
from paddleocr import PaddleOCR, draw_ocr
# 模型路径下必须含有model和params文件
ocr = PaddleOCR(det_model_dir='models/li_det_infer', use_gpu=False)
dir_path="D:/iiiiii/img_50/"
dir_path2="D:/result02/"
files = os.listdir(dir_path)
print(files)
for file in files :
img_path = dir_path+str(file)
result = ocr.ocr(img_path, rec=False)
fo = open(dir_path2+str(file)[:-4]+".txt", "w")
for line in result:
fo.write(str(line) + "\n")
fo.close()
from PIL import Image
image = Image.open(img_path).convert('RGB')
# boxes = [line[0] for line in result]
# txts = [line[1][0] for line in result]
# scores = [line[1][1] for line in result]
im_show = draw_ocr(image, result,txts=None, scores=None,font_path='D:/paddle_pp/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
from paddleocr import PaddleOCR, draw_ocr
# 模型路径下必须含有model和params文件
ocr = PaddleOCR(rec_model_dir='models/li_rec_infer',
use_angle_cls=True,
use_gpu=False)
img_path = 'imgs/xxxxxxxx.png'
result = ocr.ocr(img_path, cls=True)
for line in result:
print(line)
# 显示结果
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image,
boxes,
txts,
scores,
font_path='/path/to/PaddleOCR/doc/fonts/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
from paddleocr import PaddleOCR, draw_ocr
ocr = PaddleOCR(use_angle_cls=True, lang="ch")
# your own path
img_path = "../img/test2.jpg"
result = ocr.ocr(img_path, cls=True)
for line in result:
print(line)
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result2.jpg')