diff --git a/README.md b/README.md index 114e7f6033d7d7d092ab09f493c32ebd56acac5f..4eb3edda162d641722830e35b2a165f9cc891182 100644 --- a/README.md +++ b/README.md @@ -11,7 +11,7 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库，助力 ## **超轻量级中文OCR体验** -![](./doc/imgs_draw/11.jpg) +![](doc/imgs_results/11.jpg) 上图是超轻量级中文OCR模型效果展示，更多效果图请见文末[效果展示](#效果展示)。 @@ -97,14 +97,14 @@ PaddleOCR文本识别算法的训练和使用请参考文档教程中[文本识 ## 效果展示 -![](./doc/imgs_draw/1.jpg) -![](./doc/imgs_draw/7.jpg) -![](./doc/imgs_draw/12.jpg) -![](./doc/imgs_draw/4.jpg) -![](./doc/imgs_draw/6.jpg) -![](./doc/imgs_draw/9.jpg) -![](./doc/imgs_draw/16.png) -![](./doc/imgs_draw/22.jpg) +![](doc/imgs_results/1.jpg) +![](doc/imgs_results/7.jpg) +![](doc/imgs_results/12.jpg) +![](doc/imgs_results/4.jpg) +![](doc/imgs_results/6.jpg) +![](doc/imgs_results/9.jpg) +![](doc/imgs_results/16.png) +![](doc/imgs_results/22.jpg) ## 参考文献 diff --git a/doc/imgs/.DS_Store b/doc/imgs/.DS_Store deleted file mode 100644 index c35fcc4fd912173bc5f7b84f65bedd4d00f38c7b..0000000000000000000000000000000000000000 Binary files a/doc/imgs/.DS_Store and /dev/null differ diff --git a/doc/imgs_draw/1.jpg b/doc/imgs_results/1.jpg similarity index 100% rename from doc/imgs_draw/1.jpg rename to doc/imgs_results/1.jpg diff --git a/doc/imgs_draw/10.jpg b/doc/imgs_results/10.jpg similarity index 100% rename from doc/imgs_draw/10.jpg rename to doc/imgs_results/10.jpg diff --git a/doc/imgs_draw/11.jpg b/doc/imgs_results/11.jpg similarity index 100% rename from doc/imgs_draw/11.jpg rename to doc/imgs_results/11.jpg diff --git a/doc/imgs_draw/12.jpg b/doc/imgs_results/12.jpg similarity index 100% rename from doc/imgs_draw/12.jpg rename to doc/imgs_results/12.jpg diff --git a/doc/imgs_draw/13.png b/doc/imgs_results/13.png similarity index 100% rename from doc/imgs_draw/13.png rename to doc/imgs_results/13.png diff --git a/doc/imgs_draw/15.jpg b/doc/imgs_results/15.jpg similarity index 100% rename from doc/imgs_draw/15.jpg rename to doc/imgs_results/15.jpg diff --git a/doc/imgs_draw/16.png b/doc/imgs_results/16.png similarity index 100% rename from doc/imgs_draw/16.png rename to doc/imgs_results/16.png diff --git a/doc/imgs_draw/17.png b/doc/imgs_results/17.png similarity index 100% rename from doc/imgs_draw/17.png rename to doc/imgs_results/17.png diff --git a/doc/imgs_draw/2.jpg b/doc/imgs_results/2.jpg similarity index 100% rename from doc/imgs_draw/2.jpg rename to doc/imgs_results/2.jpg diff --git a/doc/imgs_draw/22.jpg b/doc/imgs_results/22.jpg similarity index 100% rename from doc/imgs_draw/22.jpg rename to doc/imgs_results/22.jpg diff --git a/doc/imgs_draw/3.jpg b/doc/imgs_results/3.jpg similarity index 100% rename from doc/imgs_draw/3.jpg rename to doc/imgs_results/3.jpg diff --git a/doc/imgs_draw/4.jpg b/doc/imgs_results/4.jpg similarity index 100% rename from doc/imgs_draw/4.jpg rename to doc/imgs_results/4.jpg diff --git a/doc/imgs_draw/5.jpg b/doc/imgs_results/5.jpg similarity index 100% rename from doc/imgs_draw/5.jpg rename to doc/imgs_results/5.jpg diff --git a/doc/imgs_draw/6.jpg b/doc/imgs_results/6.jpg similarity index 100% rename from doc/imgs_draw/6.jpg rename to doc/imgs_results/6.jpg diff --git a/doc/imgs_draw/7.jpg b/doc/imgs_results/7.jpg similarity index 100% rename from doc/imgs_draw/7.jpg rename to doc/imgs_results/7.jpg diff --git a/doc/imgs_draw/8.jpg b/doc/imgs_results/8.jpg similarity index 100% rename from doc/imgs_draw/8.jpg rename to doc/imgs_results/8.jpg diff --git a/doc/imgs_draw/9.jpg b/doc/imgs_results/9.jpg similarity index 100% rename from doc/imgs_draw/9.jpg rename to doc/imgs_results/9.jpg diff --git a/tools/infer/predict_system.py b/tools/infer/predict_system.py index 4907a7ccf0105b076e702e169af519b02091d654..83acdbe5edbbdc1b5cf92565ebbe787304e8d103 100755 --- a/tools/infer/predict_system.py +++ b/tools/infer/predict_system.py @@ -22,6 +22,10 @@ import copy import numpy as np import math import time +from ppocr.utils.utility import get_image_file_list +from PIL import Image +from tools.infer.utility import draw_ocr +import os class TextSystem(object): @@ -99,8 +103,9 @@ def sorted_boxes(dt_boxes): if __name__ == "__main__": args = utility.parse_args() - image_file_list = utility.get_image_file_list(args.image_dir) + image_file_list = get_image_file_list(args.image_dir) text_sys = TextSystem(args) + is_visualize = True for image_file in image_file_list: img = cv2.imread(image_file) if img is None: @@ -114,8 +119,22 @@ if __name__ == "__main__": dt_boxes_final = [] for dno in range(dt_num): text, score = rec_res[dno] - if score >= 0: + if score >= 0.5: text_str = "%s, %.3f" % (text, score) print(text_str) dt_boxes_final.append(dt_boxes[dno]) - utility.draw_text_det_res(dt_boxes_final, image_file) + + if is_visualize: + image = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB)) + boxes = dt_boxes + txts = [rec_res[i][0] for i in range(len(rec_res))] + scores = [rec_res[i][1] for i in range(len(rec_res))] + + draw_img = draw_ocr( + image, boxes, txts, scores, draw_txt=True, drop_score=0.5) + draw_img_save = "./doc/imgs_results/" + if not os.path.exists(draw_img_save): + os.makedirs(draw_img_save) + cv2.imwrite( + os.path.join(draw_img_save, os.path.basename(image_file)), + draw_img) diff --git a/tools/infer/utility.py b/tools/infer/utility.py index 7132ddfa08c4dc9b3fc60b9a5cd2e5b51336a3c9..8f2bd754201287f2eb652052253d26ebda412121 100755 --- a/tools/infer/utility.py +++ b/tools/infer/utility.py @@ -112,36 +112,70 @@ def draw_text_det_res(dt_boxes, img_path): cv2.imwrite("./output/%s" % img_name_pure, src_im) -def draw_ocr(image, boxes, txts, scores, draw_txt): +def resize_img(img, input_size=600): + """ + """ + img = np.array(img) + im_shape = img.shape + im_size_min = np.min(im_shape[0:2]) + im_size_max = np.max(im_shape[0:2]) + im_scale = float(input_size) / float(im_size_max) + im = cv2.resize(img, None, None, fx=im_scale, fy=im_scale) + return im + + +def draw_ocr(image, boxes, txts, scores, draw_txt=True, drop_score=0.5): from PIL import Image, ImageDraw, ImageFont w, h = image.size img = image.copy() draw = ImageDraw.Draw(img) - for (box, txt) in zip(boxes, txts): - + for (box, score) in zip(boxes, scores): + if score < drop_score: + continue draw.line([(box[0][0], box[0][1]), (box[1][0], box[1][1])], fill='red') draw.line([(box[1][0], box[1][1]), (box[2][0], box[2][1])], fill='red') draw.line([(box[2][0], box[2][1]), (box[3][0], box[3][1])], fill='red') draw.line([(box[3][0], box[3][1]), (box[0][0], box[0][1])], fill='red') + draw.line( + [(box[0][0] - 1, box[0][1] + 1), (box[1][0] - 1, box[1][1] + 1)], + fill='red') + draw.line( + [(box[1][0] - 1, box[1][1] + 1), (box[2][0] - 1, box[2][1] + 1)], + fill='red') + draw.line( + [(box[2][0] - 1, box[2][1] + 1), (box[3][0] - 1, box[3][1] + 1)], + fill='red') + draw.line( + [(box[3][0] - 1, box[3][1] + 1), (box[0][0] - 1, box[0][1] + 1)], + fill='red') if draw_txt: txt_color = (0, 0, 0) - - blank_img = np.ones(shape=[h, 800], dtype=np.int8) * 255 + img = np.array(resize_img(img)) + _h = img.shape[0] + blank_img = np.ones(shape=[_h, 600], dtype=np.int8) * 255 blank_img = Image.fromarray(blank_img).convert("RGB") draw_txt = ImageDraw.Draw(blank_img) - font_size = 30 - gap = 40 if h // len(txts) >= font_size else h // len(txts) - - for i, txt in enumerate(txts): + font_size = 20 + gap = 20 + title = "index text score" + font = ImageFont.truetype( + "./doc/simfang.ttf", font_size, encoding="utf-8") + + draw_txt.text((20, 0), title, txt_color, font=font) + count = 0 + for idx, txt in enumerate(txts): + if scores[idx] < drop_score: + continue font = ImageFont.truetype( - "./doc/simfang.TTF", font_size, encoding="utf-8") - new_txt = str(i) + ': ' + txt + ' ' + str(scores[i]) - draw_txt.text((20, gap * (i + 1)), new_txt, txt_color, font=font) - + "./doc/simfang.ttf", font_size, encoding="utf-8") + new_txt = str(count) + ': ' + txt + ' ' + str(scores[count]) + draw_txt.text( + (20, gap * (count + 1)), new_txt, txt_color, font=font) + count += 1 img = np.concatenate([np.array(img), np.array(blank_img)], axis=1) return img