基于CCPD的车牌号检测和识别案例

examples/license_plate_detection_and_recognition/db_r50_ccpd.yaml

@@ -0,0 +1,152 @@
+system:
+  mode: 0 # 0 for graph mode, 1 for pynative mode in MindSpore
+  distribute: False
+  amp_level: 'O0'
+  seed: 42
+  log_interval: 10
+  val_while_train: False
+  drop_overflow_update: False
+
+model:
+  type: det
+  transform: null
+  backbone:
+    name: det_resnet50
+    pretrained: False
+  neck:
+    name: DBFPN
+    out_channels: 256
+    bias: False
+  head:
+    name: DBHead
+    k: 50
+    bias: False
+    adaptive: True
+  pretrained: https://download.mindspore.cn/toolkits/mindocr/dbnet/dbnet_resnet50_synthtext-40655acb.ckpt
+
+postprocess:
+  name: DBPostprocess
+  box_type: quad          # whether to output a polygon or a box
+  binary_thresh: 0.3      # binarization threshold
+  box_thresh: 0.7         # box score threshold
+  max_candidates: 1000
+  expand_ratio: 1.5       # coefficient for expanding predictions
+
+metric:
+  name: DetMetric
+  main_indicator: f-score
+
+loss:
+  name: DBLoss
+  eps: 1.0e-6
+  l1_scale: 10
+  bce_scale: 5
+  bce_replace: bceloss
+
+scheduler:
+  scheduler: polynomial_decay
+  lr: 0.007
+  num_epochs: 1200
+  decay_rate: 0.9
+  warmup_epochs: 3
+
+optimizer:
+  opt: SGD
+  filter_bias_and_bn: false
+  momentum: 0.9
+  weight_decay: 1.0e-4
+
+# only used for mixed precision training
+loss_scaler:
+  type: dynamic
+  loss_scale: 512
+  scale_factor: 2
+  scale_window: 1000
+
+train:
+  ckpt_save_dir: './dbnet_ccpd'
+  dataset_sink_mode: True
+  dataset:
+    type: DetDataset
+    dataset_root: path/to/DBNet_DataSets
+    data_dir: train/images
+    label_file: train/train_det_gt.txt
+    sample_ratio: 1.0
+    transform_pipeline:
+      - DecodeImage:
+          img_mode: RGB
+          to_float32: False
+      - DetLabelEncode:
+      - RandomColorAdjust:
+          brightness: 0.1255  # 32.0 / 255
+          saturation: 0.5
+      - RandomHorizontalFlip:
+          p: 0.5
+      - RandomScale:
+          scale_range: [ 0.5, 3.0 ]
+          p: 1.0
+      - RandomCropWithBBox:
+          max_tries: 10
+          min_crop_ratio: 0.1
+          crop_size: [ 640, 640 ]
+          p: 1.0
+      - ValidatePolygons:
+      - ShrinkBinaryMap:
+          min_text_size: 8
+          shrink_ratio: 0.55
+      - BorderMap:
+          shrink_ratio: 0.4
+          thresh_min: 0.3
+          thresh_max: 0.7
+      - NormalizeImage:
+          bgr_to_rgb: False
+          is_hwc: True
+          mean: imagenet
+          std: imagenet
+      - ToCHWImage:
+    #  the order of the dataloader list, matching the network input and the input labels for the loss function, and optional data for debug/visualize
+    output_columns: [ 'image', 'binary_map', 'mask', 'thresh_map', 'thresh_mask' ] #'img_path']
+#    output_columns: ['image'] # for debug op performance
+    net_input_column_index: [0] # input indices for network forward func in output_columns
+    label_column_index: [1, 2, 3, 4] # input indices marked as label
+
+  loader:
+    shuffle: True
+    batch_size: 16
+    drop_remainder: True
+    num_workers: 20
+
+eval:
+  ckpt_load_path: dbnet_ccpd/best.ckpt
+  dataset_sink_mode: False
+  dataset:
+    type: DetDataset
+    dataset_root: path/to/DBNet_DataSets
+    data_dir: val/images
+    label_file: val/val_det_gt.txt
+    sample_ratio: 1.0
+    transform_pipeline:
+      - DecodeImage:
+          img_mode: RGB
+          to_float32: False
+      - DetLabelEncode:
+      - DetResize:
+          target_size: [ 1024, 1024 ] # h, w
+          keep_ratio: True
+          padding: True
+      - NormalizeImage:
+          bgr_to_rgb: False
+          is_hwc: True
+          mean: imagenet
+          std: imagenet
+      - ToCHWImage:
+    #  the order of the dataloader list, matching the network input and the labels for evaluation
+    output_columns: [ 'image', 'polys', 'ignore_tags', 'shape_list'  ]
+    net_input_column_index: [0] # input indices for network forward func in output_columns
+    label_column_index: [1, 2] # input indices marked as label
+
+  loader:
+    shuffle: False
+    batch_size: 16 # TODO: due to dynamic shape of polygons (num of boxes varies), BS has to be 1
+    drop_remainder: True
+    num_workers: 2

examples/license_plate_detection_and_recognition/generate_data.py

@@ -0,0 +1,46 @@
+import json
+import os
+
+from PIL import Image
+
+
+def read_annotations(file_path):
+    annotations = []
+    with open(file_path, "r", encoding="utf-8") as f:
+        for line in f:
+            image_name, description = line.strip().split("\t")
+            description_data = json.loads(description)
+            annotations.append((image_name, description_data))
+    return annotations
+
+
+def crop_images(annotations, source_folder, target_folder, output_txt):
+    if not os.path.exists(target_folder):
+        os.makedirs(target_folder)
+    with open(output_txt, "w", encoding="utf-8") as out_file:
+        for image_name, data in annotations:
+            image_path = os.path.join(source_folder, image_name)
+            with Image.open(image_path) as img:
+                bbox = data[0]["bbox"]
+                x1, y1 = bbox[0]
+                x2, y2 = bbox[1]
+                cropped_img = img.crop((x1, y1, x2, y2))
+                cropped_image_name = f"{image_name}"
+                cropped_img.save(os.path.join(target_folder, cropped_image_name))
+                transcription = data[0]["transcription"]
+                out_file.write(f"{cropped_image_name}\t{transcription}\n")
+
+
+def main():
+    datasets = ["train", "test", "val"]
+    for dataset in datasets:
+        annotations_file = f"path/to/DBNet_DataSets/{dataset}/{dataset}_det_gt.txt"
+        source_folder = f"path/to/DBNet_DataSets/{dataset}/images"
+        target_folder = f"path/to/SVTR_DataSets/{dataset}/"
+        output_txt = f"path/to/SVTR_DataSets/gt_{dataset}.txt"
+        annotations = read_annotations(annotations_file)
+        crop_images(annotations, source_folder, target_folder, output_txt)
+
+
+if __name__ == "__main__":
+    main()

examples/license_plate_detection_and_recognition/generate_dict.py

@@ -0,0 +1,107 @@
+provinces = [
+    "皖",
+    "沪",
+    "津",
+    "渝",
+    "冀",
+    "晋",
+    "蒙",
+    "辽",
+    "吉",
+    "黑",
+    "苏",
+    "浙",
+    "京",
+    "闽",
+    "赣",
+    "鲁",
+    "豫",
+    "鄂",
+    "湘",
+    "粤",
+    "桂",
+    "琼",
+    "川",
+    "贵",
+    "云",
+    "藏",
+    "陕",
+    "甘",
+    "青",
+    "宁",
+    "新",
+    "警",
+    "学",
+    "O",
+]
+alphabets = [
+    "A",
+    "B",
+    "C",
+    "D",
+    "E",
+    "F",
+    "G",
+    "H",
+    "J",
+    "K",
+    "L",
+    "M",
+    "N",
+    "P",
+    "Q",
+    "R",
+    "S",
+    "T",
+    "U",
+    "V",
+    "W",
+    "X",
+    "Y",
+    "Z",
+    "O",
+]
+ads = [
+    "A",
+    "B",
+    "C",
+    "D",
+    "E",
+    "F",
+    "G",
+    "H",
+    "J",
+    "K",
+    "L",
+    "M",
+    "N",
+    "P",
+    "Q",
+    "R",
+    "S",
+    "T",
+    "U",
+    "V",
+    "W",
+    "X",
+    "Y",
+    "Z",
+    "0",
+    "1",
+    "2",
+    "3",
+    "4",
+    "5",
+    "6",
+    "7",
+    "8",
+    "9",
+    "O",
+]
+
+unique_characters = set(provinces + alphabets + ads)
+unique_dict = {char: index for index, char in enumerate(sorted(unique_characters))}
+with open("ccpd.txt", "w", encoding="utf-8") as file:
+    for char, index in unique_dict.items():
+        line = f"{char}:{index}\n"
+        file.write(line)