yolov7中的mosaic增强实现原理
yolov7源码链接：GitHub - WongKinYiu/yolov7: Implementation of paper - YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors
一、mosaic增强的总体思想
1、构建一个画布
s = self.img_size
img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8)2、为该画布生成随机中心点
随机中心点的取值范围为：[img_size // 2, int(1.5*img_size )]，即0.5*img_size到1.5img_size
self.mosaic_border = [-img_size // 2, -img_size // 2]
yc, xc = [int(random.uniform(-x, 2 * s + x)) for x in self.mosaic_border]3、将4张图片依次填入画布
左上
x1a（纵坐标同理）的取值有两种情况，
1、xc-w<=0，则x1a=0，x1b=w-xc，padw=xc-w（padw用于修正label）
2、xc-w>0，则x1a=xc-w，x1b=0，padw=xc-w（这种情况意味着小图片可以完全放入画布的左上方区域）
x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc
x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
padw = x1a - x1b
padh = y1a - y1b右上
x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc
x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
padw = x1a - x1b
padh = y1a - y1b左下
x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h)
x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h)
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
padw = x1a - x1b
padh = y1a - y1b右下
x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h)
x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h)
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
padw = x1a - x1b
padh = y1a - y1b4、修正label
labels, segments = self.labels[index].copy(), self.segments[index].copy()
if labels.size:
labels[:, 1:] = xywhn2xyxy(labels[:, 1:], w, h, padw, padh) # normalized xywh to pixel xyxy format
segments = [xyn2xy(x, w, h, padw, padh) for x in segments]
labels4.append(labels)
segments4.extend(segments)labels[:, 1:]表示目标的坐标，格式是xywh，且是归一化后的坐标
利用xywhn2xyxy将以上坐标转换至w、h尺度上的坐标，并修正了坐标（根据padw和padh），坐标格式变成x1y1x2y2
def xywhn2xyxy(x, w=640, h=640, padw=0, padh=0):
# Convert nx4 boxes from [x, y, w, h] normalized to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
y[:, 0] = w * (x[:, 0] - x[:, 2] / 2) + padw # top left x
y[:, 1] = h * (x[:, 1] - x[:, 3] / 2) + padh # top left y
y[:, 2] = w * (x[:, 0] + x[:, 2] / 2) + padw # bottom right x
y[:, 3] = h * (x[:, 1] + x[:, 3] / 2) + padh # bottom right y
return ysegments同理
def xyn2xy(x, w=640, h=640, padw=0, padh=0):
# Convert normalized segments into pixel segments, shape (n,2)
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
y[:, 0] = w * x[:, 0] + padw # top left x
y[:, 1] = h * x[:, 1] + padh # top left y
return y二、mosaic增强函数具体实现
yolov7的mosaic增强函数如下，
def load_mosaic(self, index):
# loads images in a 4-mosaic
labels4, segments4 = [], []
s = self.img_size
yc, xc = [int(random.uniform(-x, 2 * s + x)) for x in self.mosaic_border]
indices = [index] + random.choices(self.indices, k=3) # 3 additional image indices
for i, index in enumerate(indices):
# Load image
img, _, (h, w) = load_image(self, index)
# place img in img4
if i == 0: # top left
# base image with 4 tiles
img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8)
# xmin, ymin, xmax, ymax (large image)
x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc
# xmin, ymin, xmax, ymax (small image)
x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h
elif i == 1: # top right
x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc
x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h
elif i == 2: # bottom left
x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h)
x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h)
elif i == 3: # bottom right
x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h)
x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h)
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b] # img4[ymin:ymax, xmin:xmax]
padw = x1a - x1b
padh = y1a - y1b
# Labels
labels, segments = self.labels[index].copy(), self.segments[index].copy()
if labels.size:
labels[:, 1:] = xywhn2xyxy(labels[:, 1:], w, h, padw, padh) # normalized xywh to pixel xyxy format
segments = [xyn2xy(x, w, h, padw, padh) for x in segments]
labels4.append(labels)
segments4.extend(segments)
# Concat/clip labels
labels4 = np.concatenate(labels4, 0)
for x in (labels4[:, 1:], *segments4):
np.clip(x, 0, 2 * s, out=x) # clip when using random_perspective()
# Augment
img4, labels4, segments4 = copy_paste(img4, labels4, segments4, probability=self.hyp['copy_paste'])
img4, labels4 = random_perspective(img4, labels4, segments4,
degrees=self.hyp['degrees'],
translate=self.hyp['translate'],
scale=self.hyp['scale'],
shear=self.hyp['shear'],
perspective=self.hyp['perspective'],
border=self.mosaic_border) # border to remove
return img4, labels4三、效果展示
演示代码如下，
import random
import numpy as np
import cv2
def load_mosaic(imgs):
s = 320
mosaic_border = [-160, -160]
yc, xc = [int(random.uniform(-x, 2 * s + x)) for x in mosaic_border]
for i in range(4):
img = imgs[i]
h, w = img.shape[0:2]
if i == 0: # top left
img4 = np.full((s * 2, s * 2, 3), 114, dtype=np.uint8)
x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc
x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h
elif i == 1: # top right
x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc
x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h
elif i == 2: # bottom left
x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h)
x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h)
elif i == 3: # bottom right
x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h)
x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h)
img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
cv2.imwrite("/home/projects/yolov7/debug/img_concat.jpg", img4)
img1 = cv2.resize(cv2.imread("/home/projects/yolov7/debug/img1.jpg"), [320, 320])
img2 = cv2.resize(cv2.imread("/home/projects/yolov7/debug/img2.jpg"), [320, 320])
img3 = cv2.resize(cv2.imread("/home/projects/yolov7/debug/img3.jpg"), [320, 320])
img4 = cv2.resize(cv2.imread("/home/projects/yolov7/debug/img4.jpg"), [320, 320])
imgs = [img1, img2, img3, img4]
load_mosaic(imgs=imgs)mosaic增强前的图片
mosaic增强后的图片
