yolov5 区域识别具体实现
YOLOv5(You Only Look Once version 5)是目标检测算法 YOLO(You Only Look Once)的最新版本,由美国加州大学戴维斯分校研究人员 Alexey Bochkovskiy, Chien-Yao Wang 和 Hong-Yuan Mark Liao 开发。它使用深度学习技术,结合卷积神经网络和目标检测技术,能够在实时视频中识别和定位多个目标,包括人、车、动物等。相比前几个版本,YOLOv5的性能提升了很多,速度更快,精度更高,并且支持多种架构。
在YOLOv5中,区域识别的具体实现可以通过以下步骤完成:
1.数据预处理:首先,需要对输入的图像进行预处理,将其缩放到固定的尺寸大小,并将像素值归一化到0到1的范围内。此外,还需要进行数据增强的操作,包括随机裁剪、旋转、翻转等,以增加训练数据的多样性和鲁棒性。
2.网络结构:YOLOv5采用了一种基于骨干网络的轻量级网络结构,即CSPDarknet。该网络结构由多个连续的CSP块组成,每个CSP块由一个卷积层和多个残差块组成,其中残差块内部也包含了CSP结构。这种设计可以使得模型在保证高精度的情况下,具有更高的速度和更小的模型体积。
3.损失函数:YOLOv5采用了一种新的损失函数——CIoU loss,用于优化检测框的预测结果。具体来说,CIoU loss包括目标置信度损失(objectness loss)、分类损失(classification loss)和位置损失(box regression loss),其中位置损失部分还包括了一些特殊的技巧,如GIoU、DIoU等。
4.预测与后处理:在训练好模型后,需要对测试图像进行预测,并对预测结果进行后处理,包括非极大值抑制(NMS)、置信度阈值过滤、目标尺寸还原等操作,最终得到检测结果。
以下是使用yolov5进行目标检测和区域识别的大致代码流程:
import cv2
import torch
import numpy as np
from yolov5 import detect
from PIL import Image
载入模型
model = torch.hub.load(‘ultralytics/yolov5’, ‘yolov5s’)
定义类别信息
class_names = [‘person’, ‘car’, ‘bicycle’, ‘motorcycle’, ‘truck’, ‘bus’, ‘train’]
读取图片
image = cv2.imread(‘test.jpg’)
进行目标检测
results = model(image)
提取预测结果
pred_boxes = results.xyxy[0].numpy()
pred_classes = results.pred[0].numpy()
遍历每个预测结果,进行区域识别
for i in range(len(pred_boxes)):
box = pred_boxes[i]
label = class_names[int(pred_classes[i])]
xmin, ymin, xmax, ymax = box
# 提取目标区域
roi = image[int(ymin):int(ymax), int(xmin):int(xmax)]
# 进行区域识别,以车辆为例
if label in [‘car’, ‘truck’, ‘bus’]:
# 转换为PIL格式
roi = Image.fromarray(cv2.cvtColor(roi, cv2.COLOR_BGR2RGB))
# 进行分类,这里需要自己定义车辆分类器
class_id = classify_car(roi)
# 输出结果
print(‘Found a {} with class id {} in region ({}, {}, {}, {})’.format(label, class_id, xmin, ymin, xmax, ymax))
其中, classify_car() 函数需要自己定义。可以使用机器学习或深度学习的方法对车辆进行分类。常用的方法包括传统的SVM、KNN和最近比较火的CNN模型等
## 以下根据yolov5原detect.py文件做的更改
对需要识别的区域添加mask蒙版,完整代码如下:
```python
import argparse
import base64
import os
import platform
import sys
import datetime
import time
from pathlib import Path
import numpy as np
import pymysql as ml
import torch
from models.experimental import attempt_load
from utils.datasets import letterbox
FILE = Path(__file__).resolve()
ROOT = FILE.parents[0] # YOLOv5 root directory
if str(ROOT) not in sys.path:
sys.path.append(str(ROOT)) # add ROOT to PATH
ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative
from models.common import DetectMultiBackend
from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams
from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2,
increment_path, non_max_suppression, print_args, scale_boxes, strip_optimizer, xyxy2xywh,
set_logging, apply_classifier)
from utils.plots import Annotator, colors, save_one_box
from utils.torch_utils import select_device, smart_inference_mode
@smart_inference_mode()
def run(
weights=ROOT / 'yolov5s.pt', # model path or triton URL
source=ROOT / 'data/images', # file/dir/URL/glob/screen/0(webcam)
data=ROOT / 'data/coco128.yaml', # dataset.yaml path
imgsz=(480, 640), # inference size (height, width)
conf_thres=0.45, # confidence threshold
iou_thres=0.65, # NMS IOU threshold
max_det=1000, # maximum detections per image
device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu
view_img=False, # show results
save_txt=False, # save results to *.txt
save_conf=False, # save confidences in --save-txt labels
save_crop=False, # save cropped prediction boxes
nosave=False, # do not save images/videos
classes=None, # filter by class: --class 0, or --class 0 2 3
agnostic_nms=False, # class-agnostic NMS
augment=False, # augmented inference
visualize=False, # visualize features
update=False, # update all models
project=ROOT / 'runs/detect', # save results to project/name
name='exp', # save results to project/name
exist_ok=False, # existing project/name ok, do not increment
line_thickness=3, # bounding box thickness (pixels)
hide_labels=False, # hide labels
hide_conf=False, # hide confidences
half=False, # use FP16 half-precision inference
dnn=False, # use OpenCV DNN for ONNX inference
vid_stride=1, # video frame-rate stride
):
source = str(source)
save_img = not nosave and not source.endswith('.txt') # save inference images
is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS)
is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://'))
webcam = source.isnumeric() or source.endswith('.streams') or (is_url and not is_file)
screenshot = source.lower().startswith('screen')
if is_url and is_file:
source = check_file(source) # download
# Directories
save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run
(save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir
# Load model
device = select_device(device)
model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half)
stride, names, pt = model.stride, model.names, model.pt
imgsz = check_img_size(imgsz, s=stride) # check image size
# Dataloader
bs = 1 # batch_size
# 判断是摄像机来源还是图片
is_camera = False
if webcam:
is_camera = True
view_img = check_imshow(warn=True)
dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride)
bs = len(dataset)
elif screenshot:
dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt)
else:
dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride)
vid_path, vid_writer = [None] * bs, [None] * bs
# Run inference
model.warmup(imgsz=(1 if pt or model.triton else bs, 3, *imgsz)) # warmup
seen, windows, dt = 0, [], (Profile(), Profile(), Profile())
for path, im, im0s, vid_cap, s in dataset:
# =====================画mask蒙版============================================================
# mask for certain region
# 1,2,3,4 分别对应左上,右上,右下,左下四个点
hl1 = 4.8 / 10 # 监测区域高度距离图片顶部比例(左上)
hl2 = 4.6 / 10 # 监测区域高度距离图片顶部比例(右上)
hl3 = 9.9 / 10 # 监测区域高度距离图片顶部比例(右下)
hl4 = 9.9 / 10 # 监测区域高度距离图片顶部比例(左下)
hl5 = 6.9 / 10 # 监测区域高度距离图片顶部比例(左下)
wl1 = 5.4 / 10 # 监测区域高度距离图片左部比例(上水平距左)
wl2 = 9.95 / 10 # 监测区域高度距离图片左部比例(上水平距右)
wl3 = 9.9 / 10 # 监测区域高度距离图片左部比例(下水平距右)
wl4 = 4.8 / 10 # 监测区域高度距离图片左部比例(下水平距左)
wl5 = 3.8 / 10 # 监测区域高度距离图片左部比例(下水平距左)
if webcam:
for b in range(0, im.shape[0]):
mask = np.zeros([im[b].shape[1], im[b].shape[2]], dtype=np.uint8)
# mask[round(img[b].shape[1] * hl1):img[b].shape[1], round(img[b].shape[2] * wl1):img[b].shape[2]] = 255
pts = np.array([[int(im[b].shape[2] * wl1), int(im[b].shape[1] * hl1)], # pts1
[int(im[b].shape[2] * wl2), int(im[b].shape[1] * hl2)], # pts2
[int(im[b].shape[2] * wl3), int(im[b].shape[1] * hl3)], # pts3
[int(im[b].shape[2] * wl4), int(im[b].shape[1] * hl4)],
[int(im[b].shape[2] * wl5), int(im[b].shape[1] * hl5)]
], np.int32)
mask = cv2.fillPoly(mask, [pts], (255, 255, 255))
imgc = im[b].transpose((1, 2, 0))
imgc = cv2.add(imgc, np.zeros(np.shape(imgc), dtype=np.uint8), mask=mask)
# cv2.imshow('1',imgc)
im[b] = imgc.transpose((2, 0, 1))
else:
mask = np.zeros([im.shape[1], im.shape[2]], dtype=np.uint8)
pts = np.array([[int(im.shape[2] * wl1), int(im.shape[1] * hl1)], # pts1
[int(im.shape[2] * wl2), int(im.shape[1] * hl2)], # pts2
[int(im.shape[2] * wl3), int(im.shape[1] * hl3)], # pts3
[int(im.shape[2] * wl4), int(im.shape[1] * hl4)],
[int(im.shape[2] * wl5), int(im.shape[1] * hl5)]
], np.int32)
mask = cv2.fillPoly(mask, [pts], (255, 255, 255))
img = im.transpose((1, 2, 0))
img = cv2.add(img, np.zeros(np.shape(img), dtype=np.uint8), mask=mask)
img = img.transpose((2, 0, 1))
# =====================画mask蒙版============================================================
with dt[0]:
im = torch.from_numpy(im).to(model.device)
im = im.half() if model.fp16 else im.float() # uint8 to fp16/32
im /= 255 # 0 - 255 to 0.0 - 1.0
if len(im.shape) == 3:
im = im[None] # expand for batch dim
# Inference
with dt[1]:
visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False
pred = model(im, augment=augment, visualize=visualize)
# NMS
with dt[2]:
pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det)
# 置信度 为设置阈值
confidence = conf_thres
# Second-stage classifier (optional)
# pred = utils.general.apply_classifier(pred, classifier_model, im, im0s)
# Process predictions
for i, det in enumerate(pred): # per image
seen += 1
if webcam: # batch_size >= 1
p, im0, frame = path[i], im0s[i].copy(), dataset.count
s += f'{i}: '
device_id = f'{i}' # 设备列表index
device_name = f'{p}'
device_ip = device_name.split("/")[2].split("_")[-1]
# print(device_ip)
# =====================显示mask蒙版============================================================
pts = np.array([[int(im0.shape[1] * wl1), int(im0.shape[0] * hl1)], # pts1
[int(im0.shape[1] * wl2), int(im0.shape[0] * hl2)], # pts2
[int(im0.shape[1] * wl3), int(im0.shape[0] * hl3)], # pts3
[int(im0.shape[1] * wl4), int(im0.shape[0] * hl4)],
[int(im0.shape[1] * wl5), int(im0.shape[0] * hl5)]
], np.int32) # pts4
# pts = pts.reshape((-1, 1, 2))
zeros = np.zeros((im0.shape), dtype=np.uint8)
mask = cv2.fillPoly(zeros, [pts], color=(0, 165, 255))
im0 = cv2.addWeighted(im0, 1, mask, 0.2, 0)
cv2.polylines(im0, [pts], True, (255, 255, 0), 3)
# =====================显示mask蒙版============================================================
else:
p, im0, frame = path, im0s.copy(), getattr(dataset, 'frame', 0)
# =====================显示mask蒙版============================================================
cv2.putText(im0, "Detection_Region", (int(im0.shape[1] * wl1 - 5), int(im0.shape[0] * hl1 - 5)),
cv2.FONT_HERSHEY_SIMPLEX,
1.0, (255, 255, 0), 2, cv2.LINE_AA)
pts = np.array([[int(im0.shape[1] * wl1), int(im0.shape[0] * hl1)], # pts1
[int(im0.shape[1] * wl2), int(im0.shape[0] * hl2)], # pts2
[int(im0.shape[1] * wl3), int(im0.shape[0] * hl3)], # pts3
[int(im0.shape[1] * wl4), int(im0.shape[0] * hl4)],
[int(im0.shape[1] * wl5), int(im0.shape[0] * hl5)]
], np.int32) # pts4
zeros = np.zeros((im0.shape), dtype=np.uint8)
mask = cv2.fillPoly(zeros, [pts], color=(0, 165, 255))
im0 = cv2.addWeighted(im0, 1, mask, 0.2, 0)
cv2.polylines(im0, [pts], True, (255, 255, 0), 3)
# =====================显示mask蒙版============================================================
p = Path(p) # to Path
save_path = str(save_dir / p.name) # im.jpg
txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # im.txt
####################################保存实时检测图片路径##########################################
pic_dir = str(save_dir) + '/pic'
if not os.path.exists(pic_dir):
os.makedirs(pic_dir)
pic_path = pic_dir + '\\' + str(p.stem) + ('' if dataset.mode == 'image' else f'_{frame}')
##############################################################################################
s += '%gx%g ' % im.shape[2:] # print string
img_size = '%gx%g ' % im.shape[2:] # 图片尺寸
count = 0
class_name = ''
pic_str = ''
gn = torch.tensor(im0.shape)[[1, 0, 1, 0]] # normalization gain whwh
imc = im0.copy() if save_crop else im0 # for save_crop
annotator = Annotator(im0, line_width=line_thickness, example=str(names))
if len(det):
# Rescale boxes from img_size to im0 size
det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round()
# Print results
for c in det[:, 5].unique():
n = (det[:, 5] == c).sum() # detections per class
count = f"{n}"
class_name = 'invade'
s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string
# Write results
for *xyxy, conf, cls in reversed(det):
if save_txt: # Write to
xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh
line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format
with open(f'{txt_path}.txt', 'a') as f:
f.write(names[int(cls)] + '\t')
f.write(('%g ' * len(line)).rstrip() % line + '\n')
if save_img or save_crop or view_img: # Add bbox to image
c = int(cls) # integer class
label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}')
annotator.box_label(xyxy, label, color=colors(c, True))
##############################只保存含目标的实时检测图片#################################
if is_camera:
pic = (int(xyxy[0].item()) + int(xyxy[2].item())) / 2
if pic != 0:
cv2.imwrite(pic_path + f'{p.stem}.jpg', im0)
else:
im1 = cv2.imread('no.jpg', 1)
cv2.imwrite(pic_path + f'{p.stem}.jpg', im1)
#####################################################################################
if save_crop:
save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True)
# Stream results
im0 = annotator.result()
if view_img:
if platform.system() == 'Linux' and p not in windows:
windows.append(p)
cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux)
cv2.resizeWindow(str(p), im0.shape[1], im0.shape[0])
# cv2.imshow(str(p), im0)
cv2.waitKey(1) # 1 millisecond
# Save results (image with detections)
if save_img:
if dataset.mode == 'image':
cv2.imwrite(save_path, im0)
else: # 'video' or 'stream'
if vid_path[i] != save_path: # new video
vid_path[i] = save_path
if isinstance(vid_writer[i], cv2.VideoWriter):
vid_writer[i].release() # release previous video writer
if vid_cap: # video
fps = vid_cap.get(cv2.CAP_PROP_FPS)
w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
else: # stream
fps, w, h = 30, im0.shape[1], im0.shape[0]
save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos
vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h))
vid_writer[i].write(im0)
time.sleep(3)
# Print time (inference-only)
LOGGER.info(f"{s}{'' if len(det) else '(no detections), '}{dt[1].dt * 1E3:.1f}ms")
# Print results
t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image
LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t)
if save_txt or save_img:
s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else ''
LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}")
if update:
strip_optimizer(weights[0]) # update model (to fix SourceChangeWarning)
def parse_opt():
parser = argparse.ArgumentParser()
parser.add_argument('--weights', nargs='+', type=str, default='weights/best.pt', help='model path or triton URL')
parser.add_argument('--source', type=str, default='file.streams', help='file/dir/URL/glob/screen/0(webcam)')
parser.add_argument('--data', type=str, default=ROOT / 'data/hat.yaml', help='(optional) dataset.yaml path')
parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w')
parser.add_argument('--conf-thres', type=float, default=0.8, help='confidence threshold')
parser.add_argument('--iou-thres', type=float, default=0.45, help='NMS IoU threshold')
parser.add_argument('--max-det', type=int, default=1000, help='maximum detections per image')
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
parser.add_argument('--view-img', action='store_true', help='show results')
parser.add_argument('--save-txt', action='store_true', help='save results to *.txt')
parser.add_argument('--save-conf', action='store_true', default=True, help='save confidences in --save-txt labels')
parser.add_argument('--save-crop', action='store_true', help='save cropped prediction boxes')
parser.add_argument('--nosave', action='store_true', help='do not save images/videos')
parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --classes 0, or --classes 0 2 3')
parser.add_argument('--agnostic-nms', action='store_true', help='class-agnostic NMS')
parser.add_argument('--augment', action='store_true', help='augmented inference')
parser.add_argument('--visualize', action='store_true', help='visualize features')
parser.add_argument('--update', action='store_true', help='update all models')
parser.add_argument('--project', default=ROOT / 'runs/detect', help='save results to project/name')
parser.add_argument('--name', default='exp', help='save results to project/name')
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
parser.add_argument('--line-thickness', default=3, type=int, help='bounding box thickness (pixels)')
parser.add_argument('--hide-labels', default=False, action='store_true', help='hide labels')
parser.add_argument('--hide-conf', default=False, action='store_true', help='hide confidences')
parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference')
parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference')
parser.add_argument('--vid-stride', type=int, default=1, help='video frame-rate stride')
opt = parser.parse_args()
opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand
print_args(vars(opt))
return opt
def main(opt):
check_requirements(exclude=('tensorboard', 'thop'))
run(**vars(opt))
if __name__ == '__main__':
opt = parse_opt()
main(opt)
效果展示
![在这里插入图片描述](https://img-blog.csdnimg.cn/f4d9d76b26434bb180f28e03281d9858.png#pic_center)