【trick 6】mosaic数据增强

定义

利用了四张图片，对四张图片进行拼接，每一张图片都有其对应的框框，将四张图片拼接之后就获得一张新的图片，同时也获得这张图片对应的框框，然后我们将这样一张新的图片传入到神经网络当中去学习，相当于一下子传入四张图片进行学习了。论文中说这极大丰富了检测物体的背景！且在标准化BN计算的时候一下子会计算四张图片的数据！

PyTorch实现

def load_mosaic(self, index):
    """
    将四张图片拼接在一张马赛克图像中
    :param self:
    :param index: 需要获取的图像索引
    :return:
    """
    # loads images in a mosaic

    labels4 = []  # 拼接图像的label信息
    s = self.img_size
    # 随机初始化拼接图像的中心点坐标
    xc, yc = [int(random.uniform(s * 0.5, s * 1.5)) for _ in range(2)]  # mosaic center x, y
    # 从dataset中随机寻找三张图像进行拼接
    indices = [index] + [random.randint(0, len(self.labels) - 1) for _ in range(3)]  # 3 additional image indices
    # 遍历四张图像进行拼接 4张不同大小的图像 => 1张[1472, 1472, 3]的图像
    for i, index in enumerate(indices):
        # load image
        img, _, (h, w) = load_image(self, index)

        # place img in img4
        if i == 0:  # top left
            # 创建马赛克图像 [1472, 1472, 3]
            img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8)  # base image with 4 tiles
            # 计算马赛克图像中的坐标信息(将图像填充到马赛克图像中)
            x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc  # xmin, ymin, xmax, ymax (large image)
            # 计算截取的图像区域信息(以xc,yc为第一张图像的右下角坐标填充到马赛克图像中，丢弃越界的区域)
            x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h  # xmin, ymin, xmax, ymax (small image)
        elif i == 1:  # top right
            # 计算马赛克图像中的坐标信息(将图像填充到马赛克图像中)
            x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc
            # 计算截取的图像区域信息(以xc,yc为第二张图像的左下角坐标填充到马赛克图像中，丢弃越界的区域)
            x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h
        elif i == 2:  # bottom left
            # 计算马赛克图像中的坐标信息(将图像填充到马赛克图像中)
            x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h)
            # 计算截取的图像区域信息(以xc,yc为第三张图像的右上角坐标填充到马赛克图像中，丢弃越界的区域)
            x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, max(xc, w), min(y2a - y1a, h)
        elif i == 3:  # bottom right
            # 计算马赛克图像中的坐标信息(将图像填充到马赛克图像中)
            x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h)
            # 计算截取的图像区域信息(以xc,yc为第四张图像的左上角坐标填充到马赛克图像中，丢弃越界的区域)
            x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h)

        # 将截取的图像区域填充到马赛克图像的相应位置
        img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]  # img4[ymin:ymax, xmin:xmax]
        # 计算pad(图像边界与马赛克边界的距离，越界的情况为负值)
        padw = x1a - x1b
        padh = y1a - y1b

        # Labels 获取对应拼接图像的labels信息
        x = self.labels[index]
        labels = x.copy()  # 深拷贝，防止修改原数据
        if x.size > 0:  # Normalized xywh to pixel xyxy format
            # 计算标注数据在马赛克图像中的
            labels[:, 1] = w * (x[:, 1] - x[:, 3] / 2) + padw
            labels[:, 2] = h * (x[:, 2] - x[:, 4] / 2) + padh
            labels[:, 3] = w * (x[:, 1] + x[:, 3] / 2) + padw
            labels[:, 4] = h * (x[:, 2] + x[:, 4] / 2) + padh
        labels4.append(labels)

    # Concat/clip labels 把labels4（[(3, 5), (3, 5), (1, 5), (1, 5)] => (8, 5)）压缩到一起
    if len(labels4):
        labels4 = np.concatenate(labels4, 0)
        # np.clip(labels4[:, 1:] - s / 2, 0, s, out=labels4[:, 1:])  # use with center crop
        np.clip(labels4[:, 1:], 0, 2 * s, out=labels4[:, 1:])  # use with random_affine 防止越界

    # affine Augment  随机仿射变换 [1472, 1472, 3] => [736, 736, 3]
    # img4 = img4[s // 2: int(s * 1.5), s // 2:int(s * 1.5)]  # center crop (WARNING, requires box pruning)
    img4, labels4 = random_affine(img4, labels4,
                                  degrees=self.hyp['degrees'],
                                  translate=self.hyp['translate'],
                                  scale=self.hyp['scale'],
                                  shear=self.hyp['shear'],
                                  border=-s // 2)  # border to remove

    return img4, labels4

Reference

1. 链接: 代码.