简单使用yolov5识别滑块缺口

一、数据集准备

文件夹创建

images=>存放图片

labels=>存放标注坐标

gap.yaml =>yolov5训练配置文件

使用https://github.com/tzutalin/labelImg进行人工标注，将标注结果保存在lables文件夹

labelImg添加标注类型

注意红色位置

二、下载yolov5训练

下载地址https://github.com/ultralytics/yolov5

配置好环境，直接在命令行输入：

python train.py --img 640 --batch-size 4 --epochs 10 --data d:\MyDataset\Captcha_gap\gap.yaml --weights yolov5s.pt --nosave --cache

主要修改--batch-size，--epochs，--data

--batch-size，--epochs根据你电脑的配置修改，数值越大训练效果和时间越长，但有可能会爆内存错误，此时就得调小

--data是你准备的训练配置文件

训练结束后会在run/train下生成一个文件夹exp开头的，训练成功会生成一个weights文件夹，里面有2个文件，一个是best.pt(最优模型)一个是last.pt(最后模型)

三、进行接口推理

三种方式：

1.命令行：

python detect.py --weights runs/train/exp14/weights/best.pt --img 640 --conf 0.25 --source D:\MyDataset\Captcha_gap\images\0ca47576-3b27-4b67-97bc-bf0205fad9fd.png

2.torch.hub:

import torch

# Model
model = torch.hub.load('ultralytics/yolov5',"custom",path=r'D:\PycharmProjects\pytorch_pro\yolov5\runs\train\exp14\weights\best.pt')  # or yolov5m, yolov5l, yolov5x, custom

# Images
img = r'D:\MyDataset\Captcha_gap\images\5e6e475c-754d-4723-9412-4c048988a4d0.png'  # or file, Path, PIL, OpenCV, numpy, list

# Inference
results = model(img)

# Results
results.show()  # or .show(), .save(), .crop(), .pandas(), etc.

3.inference

import torch
import numpy as np
from models.experimental import attempt_load
from utils.general import non_max_suppression, scale_coords
from utils.torch_utils import select_device
import cv2
from random import randint
from utils.datasets import letterbox

class Detector(object):

    def __init__(self):
        self.img_size = 640
        self.threshold = 0.4
        self.max_frame = 160
        self.init_model()

    def init_model(self):

        self.weights = r'D:\PycharmProjects\pytorch_pro\yolov5\runs\train\exp14\weights\best.pt'
        self.device = '-1' if torch.cuda.is_available() else 'cpu'
        self.device = select_device(self.device)
        model = attempt_load(self.weights, map_location=self.device)
        model.to(self.device).eval()
        model.half()
        # torch.save(model, 'test.pt')
        self.m = model
        self.names = model.module.names if hasattr(
            model, 'module') else model.names
        self.colors = [
            (randint(0, 255), randint(0, 255), randint(0, 255)) for _ in self.names
        ]

    def preprocess(self, img):

        img0 = img.copy()
        img = letterbox(img, new_shape=self.img_size)[0]
        img = img[:, :, ::-1].transpose(2, 0, 1)
        img = np.ascontiguousarray(img)
        img = torch.from_numpy(img).to(self.device)
        img = img.half()  # 半精度
        img /= 255.0  # 图像归一化
        if img.ndimension() == 3:
            img = img.unsqueeze(0)

        return img0, img

    def plot_bboxes(self, image, bboxes, line_thickness=None):
        tl = line_thickness or round(
            0.002 * (image.shape[0] + image.shape[1]) / 2) + 1  # line/font thickness
        for (x1, y1, x2, y2, cls_id, conf) in bboxes:
            color = self.colors[self.names.index(cls_id)]
            c1, c2 = (x1, y1), (x2, y2)
            cv2.rectangle(image, c1, c2, color,
                          thickness=tl, lineType=cv2.LINE_AA)
            tf = max(tl - 1, 1)  # font thickness
            t_size = cv2.getTextSize(
                cls_id, 0, fontScale=tl / 3, thickness=tf)[0]
            c2 = c1[0] + t_size[0], c1[1] - t_size[1] - 3
            cv2.rectangle(image, c1, c2, color, -1, cv2.LINE_AA)  # filled
            cv2.putText(image, '{} ID-{:.2f}'.format(cls_id, conf), (c1[0], c1[1] - 2), 0, tl / 3,
                        [225, 255, 255], thickness=tf, lineType=cv2.LINE_AA)
        return image

    def detect(self, im):

        im0, img = self.preprocess(im)

        pred = self.m(img, augment=False)[0]
        pred = pred.float()
        pred = non_max_suppression(pred, self.threshold, 0.3)

        pred_boxes = []
        image_info = {}

        count = 0
        for det in pred:
            if det is not None and len(det):
                det[:, :4] = scale_coords(
                    img.shape[2:], det[:, :4], im0.shape).round()

                for *x, conf, cls_id in det:
                    lbl = self.names[int(cls_id)]
                    x1, y1 = int(x[0]), int(x[1])
                    x2, y2 = int(x[2]), int(x[3])
                    pred_boxes.append(
                        (x1, y1, x2, y2, lbl, conf))
                    count += 1
                    key = '{}-{:02}'.format(lbl, count)
                    # image_info[key] = ['{}×{}'.format(
                    #     x2-x1, y2-y1), np.round(float(conf), 3)]
                    image_info[key]={"conf":np.round(float(conf), 3),"x1":x1,"y1":y1,"x2":x2,"y2":y2}

        im = self.plot_bboxes(im, pred_boxes)
        return im, image_info


if __name__ == '__main__':
    model = Detector()
    img=cv2.imread(r'D:\MyDataset\Captcha_gap\images\5e6e475c-754d-4723-9412-4c048988a4d0.png')
    img_y, image_info = model.detect(img)
    print(img_y, image_info)

四、结果展示

{'gap-01': {'conf': 0.932, 'x1': 182, 'y1': 87, 'x2': 228, 'y2': 134}}