YOLOV5实时检测屏幕 | 您所在的位置:网站首页 › 检测代码 › YOLOV5实时检测屏幕 |
YOLOV5实时检测屏幕
目录YOLOV5实时检测屏幕思考部分先把原本的detect.py的代码贴在这里分析代码并删减不用的部分把屏幕的截图通过OpenCV进行显示写一个屏幕截图的文件用OpenCV绘制窗口并显示最终代码
注:此为笔记
目的:保留模型加载和推理部分,完成实时屏幕检测 实现思路: 1. 写一个实时截取屏幕的函数 2. 将截取的屏幕在窗口显示出来 3. 用OpenCV绘制一个窗口用来显示截取的屏幕 4. 在detect找出推理的代码,推理完成后得到中心点的xy坐标,宽高组成box 5. 在创建的OpenCV窗口用得到的推理结果绘制方框 实现效果:
做了一些包的导入,定义了一些全局变量,先保留下来,没用的最后删 向下 if __name__ == '__main__': opt = parse_opt() main(opt)从if __name__ == '__main__开始 opt = parse_opt 就是一个获取命令行参数的函数,我们并不需要,可以删 进入main函数 def main(opt): check_requirements(exclude=('tensorboard', 'thop')) run(**vars(opt))check_requirements函数检查requirements是否全都安装好了,无用,删了 进入run函数 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判断source的类型,即要要推理的源是什么,判断源是文件还是url还是webcam或者screenshot ,定义保存文件夹,我不需要保存,只需要实时检测屏幕,删除 继续向下,是加载模型的代码 # Load model device = select_device(device) model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half)得知加载模型需要几个参数,分别是weights, device=device, dnn=dnn, data=data, fp16=half 通过开始的形参可知: weights=ROOT / 'yolov5s.pt' 也就是模型的名称 device通过select_device函数得到 dnn和fp16在run函数里的参数都是FALSE故加载模型的代码可以改写成 def LoadModule(): device = select_device('') weights = 'yolov5s.pt' model = DetectMultiBackend(weights, device=device, dnn=False, fp16=False) return model继续往下读 bs = 1 # batch_size if webcam: 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这里如果是使用网络摄像头作为输入,会通过LoadStreams类加载视频流,根据图像大小和步长采样,如果使用截图作为输入,则通过LoadScreenshots加载截图,都不是则通过LoadImages类加载图片文件 这是YOLOV5提供的加载dataset的部分,我们可以添加自己的dataset,所以删掉 继续往下 # 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: 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) # Second-stage classifier (optional) # pred = utils.general.apply_classifier(pred, classifier_model, im, im0s) # Process predictionsmodel.warmup(imgsz=(1 if pt or model.triton else bs, 3, *imgsz)) 用于模型预热,传入形状为(1, 3, *imgsz)的图像进行预热操作,没用删了 seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) 未知作用,删了 for path, im, im0s, vid_cap, s in dataset: 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)上面这段for循环用于遍历数据集中的每个图像或视频帧进行推理,在循环的开头,将路径、图像、原始图像、视频捕获对象和步长传递给path, im, im0s, vid_cap, s。推理实时屏幕只需要传一张图片,所以不存在将遍历推理,所以要进行改写,改写成 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 visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False pred = model(im, augment=augment, visualize=visualize) pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det)这里是对 im 进行转换和推理,而改写的代码中没有im变量,则寻找im的来源 for path, im, im0s, vid_cap, s in dataset: im来源于dataset dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) dataset来源于LoadImages的返回值 查看LoadImages的函数返回值和返回值的来源 在dataloaders.py中可以看到 if self.transforms: im = self.transforms(im0) # transforms else: im = letterbox(im0, self.img_size, stride=self.stride, auto=self.auto)[0] # padded resize im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB im = np.ascontiguousarray(im) # contiguous return path, im, im0, self.cap, s如果transforms存在,则转换,如果transforms不存在,则调用letterbox函数对图像im0进行缩放和填充,使其符合模型要求的图像大小,将图像的通道顺序由HWC转换为CHW,将图像的通道顺序由BGR转换为RGB,将图像转换为连续的内存布局 其中需要的参数是im0, self.img_size, stride=self.stride, auto=self.auto im0则是未经处理的图片,img_size填640(因为模型的图片大小训练的是640),stride填64(默认参数为64),auto填True 则得到改写代码为 im = letterbox(img0, 640, stride=32, auto=True)[0] # padded resize im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB im = np.ascontiguousarray(im) # contiguous 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 pred = model(im, augment=False, visualize=False) pred = non_max_suppression(pred, conf_thres=conf_thres, iou_thres=iou_thres, classes=None, agnostic=False, max_det=1000)继续向下 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}: ' else: p, im0, frame = path, im0s.copy(), getattr(dataset, 'frame', 0) 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 s += '%gx%g ' % im.shape[2:] # print string 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 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 file 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(('%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)) if save_crop: save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True)这段代码将推理后的结果进行转换,转换为label format,成为人能看懂的格式,删去输出结果,留下写入结果中的,格式转换,删掉保存为txt文件,得到需要的box,然后自己写一个boxs=[],将结果append进去,方便在OpenCV中绘画识别方框,改写结果为 boxs=[] for i, det in enumerate(pred): # per image im0 = img0.copy() s = ' ' s += '%gx%g ' % im.shape[2:] # print string gn = torch.tensor(img0.shape)[[1, 0, 1, 0]] # normalization gain whwh imc = img0 # for save_crop 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 s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string # Write results for *xyxy, conf, cls in reversed(det): xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh line = (cls, *xywh) # label format box = ('%g ' * len(line)).rstrip() % line box = box.split(' ') boxs.append(box)就此完成了推理部分的删减和重写 把屏幕的截图通过OpenCV进行显示 写一个屏幕截图的文件写成 grabscreen.py # 文件名:grabscreen.py import cv2 import numpy as np import win32gui import win32print import win32ui import win32con import win32api import mss def grab_screen_win32(region): hwin = win32gui.GetDesktopWindow() left, top, x2, y2 = region width = x2 - left + 1 height = y2 - top + 1 hwindc = win32gui.GetWindowDC(hwin) srcdc = win32ui.CreateDCFromHandle(hwindc) memdc = srcdc.CreateCompatibleDC() bmp = win32ui.CreateBitmap() bmp.CreateCompatibleBitmap(srcdc, width, height) memdc.SelectObject(bmp) memdc.BitBlt((0, 0), (width, height), srcdc, (left, top), win32con.SRCCOPY) signedIntsArray = bmp.GetBitmapBits(True) img = np.fromstring(signedIntsArray, dtype='uint8') img.shape = (height, width, 4) srcdc.DeleteDC() memdc.DeleteDC() win32gui.ReleaseDC(hwin, hwindc) win32gui.DeleteObject(bmp.GetHandle()) return cv2.cvtColor(img, cv2.COLOR_BGRA2BGR)通过img0 = grab_screen_win32(region=(0, 0, 1920, 1080))来作为im的参数传入,即可让屏幕截图作为推理图片 用OpenCV绘制窗口并显示 if len(boxs): for i, det in enumerate(boxs): _, x_center, y_center, width, height = det x_center, width = re_x * float(x_center), re_x * float(width) y_center, height = re_y * float(y_center), re_y * float(height) top_left = (int(x_center - width / 2.), int(y_center - height / 2.)) bottom_right = (int(x_center + width / 2.), int(y_center + height / 2.)) color = (0, 0, 255) # RGB cv2.rectangle(img0, top_left, bottom_right, color, thickness=thickness) 和 cv2.namedWindow('windows', cv2.WINDOW_NORMAL) cv2.resizeWindow('windows', re_x // 2, re_y // 2) cv2.imshow('windows', img0) HWND = win32gui.FindWindow(None, "windows") win32gui.SetWindowPos(HWND, win32con.HWND_TOPMOST, 0, 0, 0, 0, win32con.SWP_NOMOVE | win32con.SWP_NOSIZE)结合在一起 最终代码 import torch, pynput import numpy as np import win32gui, win32con, cv2 from grabscreen import grab_screen_win32 # 本地文件 from utils.augmentations import letterbox from models.common import DetectMultiBackend from utils.torch_utils import select_device from utils.general import non_max_suppression, scale_boxes, xyxy2xywh # 可调参数 conf_thres = 0.25 iou_thres = 0.05 thickness = 2 x, y = (1920, 1080) re_x, re_y = (1920, 1080) def LoadModule(): device = select_device('') weights = 'yolov5s.pt' model = DetectMultiBackend(weights, device=device, dnn=False, fp16=False) return model model = LoadModule() while True: names = model.names img0 = grab_screen_win32(region=(0, 0, 1920, 1080)) im = letterbox(img0, 640, stride=32, auto=True)[0] # padded resize im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB im = np.ascontiguousarray(im) # contiguous 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 pred = model(im, augment=False, visualize=False) pred = non_max_suppression(pred, conf_thres=conf_thres, iou_thres=iou_thres, classes=None, agnostic=False, max_det=1000) boxs=[] for i, det in enumerate(pred): # per image im0 = img0.copy() s = ' ' s += '%gx%g ' % im.shape[2:] # print string gn = torch.tensor(img0.shape)[[1, 0, 1, 0]] # normalization gain whwh imc = img0 # for save_crop 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 s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string # Write results for *xyxy, conf, cls in reversed(det): xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh line = (cls, *xywh) # label format box = ('%g ' * len(line)).rstrip() % line box = box.split(' ') boxs.append(box) if len(boxs): for i, det in enumerate(boxs): _, x_center, y_center, width, height = det x_center, width = re_x * float(x_center), re_x * float(width) y_center, height = re_y * float(y_center), re_y * float(height) top_left = (int(x_center - width / 2.), int(y_center - height / 2.)) bottom_right = (int(x_center + width / 2.), int(y_center + height / 2.)) color = (0, 0, 255) # RGB cv2.rectangle(img0, top_left, bottom_right, color, thickness=thickness) if cv2.waitKey(1) & 0xFF == ord('q'): cv2.destroyWindow() break cv2.namedWindow('windows', cv2.WINDOW_NORMAL) cv2.resizeWindow('windows', re_x // 2, re_y // 2) cv2.imshow('windows', img0) HWND = win32gui.FindWindow(None, "windows") win32gui.SetWindowPos(HWND, win32con.HWND_TOPMOST, 0, 0, 0, 0, win32con.SWP_NOMOVE | win32con.SWP_NOSIZE)End. |
CopyRight 2018-2019 实验室设备网 版权所有 |