🦜 Object Detection 🦜

Object detection 모델을 이용한 객체 탐지 모델 만들기

SSD(Single Shot Multibox Detector) 사용

참고 Github: https://github.com/pierluigiferrari/ssd_keras
이용 모듈
- python = 3.6
- tensorflow = 1.14.0
- keras = 2.2.4

1. 데이터 수집

11종의 앵무새 데이터(csv, img)
- 11종: 고핀, 듀컵, 백색유황앵무, 붉은관유황앵무, 큰유황앵무, 스칼렛매커우, 청금강앵무, 카멜롯매커우, 홍금강앵무, 오색앵무, 회색앵무
- 각 앵무새 종 폴더에 csv, img 폴더로 구성되어 있음
img
- 300x300 이미지
csv

frame	xmin	xmax	ymin	ymax	class_id
0001_00000080.jpg	22	89	64	152	1
0001_00000047.jpg	40	124	46	160	1
0001_00000121.jpg	170	242	70	159	1

frame: image file name
xmin, xmax, ymin, ymax: 앵무새 머리 이미지의 Anchor Box pixel 위치값
class_id: 앵무새 종

2. 학습 데이터 전처리

csv, img 파일을 train, test, val로 분리

for filename in glob.iglob('모듈8데이터(SSD_앵무새)/**/*.csv', recursive=True):
    csv = pd.read_csv(filename)
    if 'train' in filename:
        train_csv = train_csv.append(csv)
        for img in glob.iglob('모듈8데이터(SSD_앵무새)/**/*.jpg', recursive=True):
            if img.split("\\")[-1] in csv['frame'].values:
                shutil.copy(img, "모듈8데이터(SSD_앵무새)/train/")
    if 'test' in filename:
        test_csv = test_csv.append(csv)
        for img in glob.iglob('모듈8데이터(SSD_앵무새)/**/*.jpg', recursive=True):
            if img.split("\\")[-1] in csv['frame'].values:
                shutil.copy(img, "모듈8데이터(SSD_앵무새)/test/")
    if 'val' in filename:
        val_csv = val_csv.append(csv)
        for img in glob.iglob('모듈8데이터(SSD_앵무새)/**/*.jpg', recursive=True):
            if img.split("\\")[-1] in csv['frame'].values:
                shutil.copy(img, "모듈8데이터(SSD_앵무새)/val/")

h5 file 생성(test, val도 동일하게 진행)

# 1: DataGenerator
train_dataset = DataGenerator(load_images_into_memory=False, hdf5_dataset_path=None)

# 2: Parse the image and label lists
train_dataset.parse_csv(images_dir='data/train/',
                        labels_filename='data/train/train.csv',
                        input_format=['image_name', 'xmin', 'xmax', 'ymin', 'ymax', 'class_id'],
                        include_classes='all')

train_dataset.create_hdf5_dataset(file_path='saved_model/dataset_train.h5',
                                  resize=False,
                                  variable_image_size=True,
                                  verbose=True)

3. 모델 선정 및 학습

build model

model: ssd 300
optimizer: Adam

K.clear_session()  # Clear previous models from memory.

model = ssd_300(image_size=(img_height, img_width, img_channels),
                n_classes=n_classes,
                mode='training',
                l2_regularization=0.0005,
                scales=scales,
                aspect_ratios_per_layer=aspect_ratios,
                two_boxes_for_ar1=two_boxes_for_ar1,
                steps=steps,
                offsets=offsets,
                clip_boxes=clip_boxes,
                variances=variances,
                normalize_coords=normalize_coords,
                subtract_mean=mean_color,
                swap_channels=swap_channels)

weights_path = './saved_model/VGG_ILSVRC_16_layers_fc_reduced.h5'
model.load_weights(weights_path, by_name=True)

adam = Adam(lr=0.0001, beta_1=0.9, beta_2=0.999, epsilon=1e-08, decay=0.0)
ssd_loss = SSDLoss(neg_pos_ratio=3, alpha=1.0)

model.compile(optimizer=adam, loss=ssd_loss.compute_loss)

training

initial_epoch = 0
final_epoch = 40
steps_per_epoch = 100

history = model.fit_generator(generator=train_generator,
                              steps_per_epoch=steps_per_epoch,
                              epochs=final_epoch,
                              callbacks=callbacks,
                              validation_data=val_generator,
                              validation_steps=ceil(val_dataset_size/batch_size),
                              initial_epoch=initial_epoch)