yolov5detect. Py comment

import argparse
import time
from pathlib import Path
import warnings
import cv2
import torch
import torch.backends.cudnn as cudnn
from numpy import random

from models.experimental import attempt_load
from utils.datasets import LoadStreams, LoadImages
from utils.general import check_img_size, check_requirements, check_imshow, non_max_suppression, apply_classifier, \
    scale_coords, xyxy2xywh, strip_optimizer, set_logging, increment_path
from utils.plots import plot_one_box
from utils.torch_utils import select_device, load_classifier, time_synchronized


def detect(save_img=False):
    #  Get output folder , Input source , The weight , Parameters and other parameters 
    source, weights, view_img, save_txt, imgsz = opt.source, opt.weights, opt.view_img, opt.save_txt, opt.img_size
    save_img = not opt.nosave and not source.endswith('.txt')  # save inference images
    webcam = source.isnumeric() or source.endswith('.txt') or source.lower().startswith(
        ('rtsp://', 'rtmp://', 'http://', 'https://'))

    #  Save the path 
    save_dir = Path(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok))  # increment run
    (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True)  # make dir

    # Initialize
    #  Get the device 
    set_logging()
    device = select_device(opt.device)
    half = device.type != 'cpu'  # half precision only supported on CUDA

    # Load model
    #  load Float32 Model , Ensure that the input picture resolution set by the user can be divided 32( If not, adjust it to be divisible and return )
    model = attempt_load(weights, map_location=device)  # load FP32 model
    stride = int(model.stride.max())  # model stride
    imgsz = check_img_size(imgsz, s=stride)  # check img_size
    if half:
        model.half()  # to FP16

    # Second-stage classifier
    #  Set the second classification , Default not to use 
    classify = False
    if classify:
        modelc = load_classifier(name='resnet101', n=2)  # initialize
        modelc.load_state_dict(torch.load('weights/resnet101.pt', map_location=device)['model']).to(device).eval()

    # Set Dataloader
    #  Different data loading methods can be set through different input sources 
    vid_path, vid_writer = None, None
    if webcam:
        view_img = check_imshow()
        cudnn.benchmark = True  # set True to speed up constant image size inference
        dataset = LoadStreams(source, img_size=imgsz, stride=stride)
    else:
        cudnn.benchmark = True
        dataset = LoadImages(source, img_size=imgsz, stride=stride)

    # Get names and colors
    #  Get category name 
    names = model.module.names if hasattr(model, 'module') else model.names
    #  Set the color of the frame 
    colors = [[random.randint(0, 255) for _ in range(3)] for _ in names]

    # Run inference
    #  Make a forward reasoning , Test whether the procedure is normal 
    if device.type != 'cpu':
        model(torch.zeros(1, 3, imgsz, imgsz).to(device).type_as(next(model.parameters())))  # run once
    t0 = time.time()
    """
     path  picture / Video path 
     img  Conduct resize+pad After the picture 
     img0  primary size picture 
     cap  When reading a picture, it is None, When reading video, it is the video source 
     """
    for path, img, im0s, vid_cap in dataset:
        img = torch.from_numpy(img).to(device)
        img = img.half() if half else img.float()  # uint8 to fp16/32
        img /= 255.0  # 0 - 255 to 0.0 - 1.0
        if img.ndimension() == 3:
            img = img.unsqueeze(0)
        """
                Forward propagation   return pred Of shape yes (1, num_boxes, 5+num_class)
               h,w The length and width of the incoming network picture , Be careful dataset Rectangular reasoning is used in detection , So here h Not necessarily equal to w
               num_boxes = h/32 * w/32 + h/16 * w/16 + h/8 * w/8
               pred[..., 0:4] Is the coordinate of the prediction frame 
                The coordinates of the prediction frame are xywh( Center point + Width length ) Format 
               pred[..., 4] by objectness Degree of confidence 
               pred[..., 5:-1] For classification results 
        """

        # Inference
        t1 = time_synchronized()
        pred = model(img, augment=opt.augment)[0]
        """
               pred: The output of forward propagation 
               conf_thres: Confidence threshold 
               iou_thres:iou threshold 
               classes: Whether to keep only specific categories 
               agnostic: Conduct nms Whether to also remove the box between different categories 
                after nms after , Forecast box format ：xywh-->xyxy( Top left, bottom right )
               pred It's a list list[torch.tensor], The length is batch_size
                every last torch.tensor Of shape by (num_boxes, 6), The content is box+conf+cls
        """
        # Apply NMS
        pred = non_max_suppression(pred, opt.conf_thres, opt.iou_thres, classes=opt.classes, agnostic=opt.agnostic_nms)
        t2 = time_synchronized()

        # Apply Classifier
        #  Add secondary classification , Default not to use 
        if classify:
            pred = apply_classifier(pred, modelc, img, im0s)

        # Process detections
        #  Process each picture 
        for i, det in enumerate(pred):  # detections per image
            if webcam:  # batch_size >= 1
                p, s, im0, frame = path[i], '%g: ' % i, im0s[i].copy(), dataset.count
            else:
                p, s, im0, frame = path, '', im0s, getattr(dataset, 'frame', 0)
            #  Set save picture / The path of the video 
            p = Path(p)  # to Path
            save_path = str(save_dir / p.name)  # img.jpg
            #  Set the coordinates of the save box txt Path to file 
            txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}')  # img.txt
            #  Set print information ( Picture length and width )
            s += '%gx%g ' % img.shape[2:]  # print string
            gn = torch.tensor(im0.shape)[[1, 0, 1, 0]]  # normalization gain whwh
            if len(det):
                # Rescale boxes from img_size to im0 size
                #  Adjust the coordinates of the prediction box ： be based on resize+pad The coordinates of the picture --> Based on the original size The coordinates of the picture 
                #  At this time, the coordinate format is xyxy
                det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()

                # Print results
                #  Print the number of categories detected 
                for c in det[:, -1].unique():
                    n = (det[:, -1] == c).sum()  # detections per class
                    s += f"{n} {names[int(c)]}{'s' * (n > 1)}, "  # add to string

                # Write results
                #  Save forecast results 
                for *xyxy, conf, cls in reversed(det):
                    if save_txt:  # Write to file
                        #  take xyxy( top left corner + The lower right corner ) Format to xywh( Center point + Width length ) Format , And divide w,h Normalization , Convert to a list and save 
                        xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist()  # normalized xywh
                        line = (cls, *xywh, conf) if opt.save_conf else (cls, *xywh)  # label format
                        with open(txt_path + '.txt', 'a') as f:
                            f.write(('%g ' * len(line)).rstrip() % line + '\n')

                    #  Draw a frame on the original picture 
                    if save_img or view_img:  # Add bbox to image
                        label = f'{names[int(cls)]} {conf:.2f}'
                        plot_one_box(xyxy, im0, label=label, color=colors[int(cls)], line_thickness=3)

            # Print time (inference + NMS)
            #  Print forward propagation +nms Time 
            print(f'{s}Done. ({t2 - t1:.3f}s)')

            # Stream results
            #  If set, display , be show picture / video 
            if view_img:
                cv2.imshow(str(p), im0)
                cv2.waitKey(1)  # 1 millisecond

            # Save results (image with detections)
            #  Set save picture / video 
            if save_img:
                if dataset.mode == 'image':
                    cv2.imwrite(save_path, im0)
                else:  # 'video' or 'stream'
                    if vid_path != save_path:  # new video
                        vid_path = save_path
                        if isinstance(vid_writer, cv2.VideoWriter):
                            vid_writer.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 += '.mp4'
                        vid_writer = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h))
                    vid_writer.write(im0)

    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 ''
        print(f"Results saved to {save_dir}{s}")

    print(f'Done. ({time.time() - t0:.3f}s)')


if __name__ == '__main__':# Parameter resolver 
    parser = argparse.ArgumentParser()
    parser.add_argument('--weights', nargs='+', type=str, default='yolov5l.pt', help='model.pt path(s)')
    parser.add_argument('--source', type=str, default='data/images', help='source')  # file/folder, 0 for webcam
    parser.add_argument('--img-size', type=int, default=640, help='inference size (pixels)')
    parser.add_argument('--conf-thres', type=float, default=0.5, help='object confidence threshold')# Degree of confidence 
    parser.add_argument('--iou-thres', type=float, default=0.20, help='IOU threshold for NMS')
    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='display results')
    parser.add_argument('--save-txt', action='store_true', help='save results to *.txt')
    parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels')
    parser.add_argument('--nosave', action='store_true', help='do not save images/videos')
    parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --class 0, or --class 0 2 3')# Multiple tags 
    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('--update', action='store_true', help='update all models')
    parser.add_argument('--project', default='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')# by true No generation expxx
    opt = parser.parse_args()
    print(opt)
    check_requirements(exclude=('pycocotools', 'thop'))

    with torch.no_grad():
        if opt.update:  # update all models (to fix SourceChangeWarning)
            for opt.weights in ['yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt']:
                detect()
                strip_optimizer(opt.weights)
        else:
            detect()