YOLOv5 visual interface

Pyside6 visual interface

Install Pyside6

Activate the previous virtual environmentyolov5

Enter the following command in the terminal of this environment

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple pyside6

Enter where pythonthe path to find the Python you are currently using

Find the designer.exe file in this path (/Lib/site-packages/PySide6/designer.exe), and then send it to the desktop to create a shortcut

UI design

1. Open designer
2. Select Main Window

3. Remove menu bar

4. Drag two labels and a line in

5. Drag two buttons into them and name them

6. Adjust the interface, fill in the files in the two labels, and make the interface smaller

7. Align to the center and check scaledContents

8. Element name changes

9. Ctrl+S save, save to the file of the previous training data set

Set up the Pyside6-uic tool

1. In the Python path of the current virtual environment, open Scriptsthe folder, find it pyside6-uic.exe, and remember the current path

2. Create tools

名称：PyUic(可自己定义)
程序：E:\kaifa\Anaconda3\envs\yolov5\Scripts\pyside6-uic.exe
实参：$FileName$ -o $FileNameWithoutExtension$.py
工作目录： $FileDir$

3. use this tool

First open the ui file just generated in pycharm

If it cannot be opened by double-clicking it, drag the file to the right to open it.

Then under [external tools] under the [Tools] tab, click [PyUic] (this name was designed by you just now)

If a .py file with the same file name is generated, the setting is successful.

Write a running program

Create template file base_ui.py

import sys
import torch
from PySide6.QtGui import QPixmap, QImage
from PySide6.QtWidgets import QMainWindow, QApplication, QFileDialog

from main_window import Ui_MainWindow #main_windows是刚才生成的ui对用的Python文件名

def convert2QImage(img):
    height, width, channel = img.shape
    return QImage(img, width, height, width * channel, QImage.Format_RGB888)

class MainWindow(QMainWindow, Ui_MainWindow):
    def __init__(self):
        super(MainWindow, self).__init__()
        self.setupUi(self)

if __name__ == "__main__":
    app = QApplication(sys.argv)

    window = MainWindow()
    window.show()

    app.exec()

Compile and run, and a pop-up window will indicate success.

Write code for internal specific functions

Explanation video: YOLOv5 Pyside6 visual interface

import sys
import cv2
import torch
from PySide6.QtGui import QPixmap, QImage
from PySide6.QtWidgets import QMainWindow, QApplication, QFileDialog
from PySide6.QtCore import QTimer

from main_window import Ui_MainWindow  # main_windows是刚才生成的ui对用的Python文件名


def convert2QImage(img):
    height, width, channel = img.shape
    return QImage(img, width, height, width * channel, QImage.Format_RGB888)


class MainWindow(QMainWindow, Ui_MainWindow):
    def __init__(self):
        super(MainWindow, self).__init__()
        self.setupUi(self)
        self.model = torch.hub.load("./", "custom", path="runs/train/exp/weights/best.pt", source="local")
        self.video = None
        self.timer = QTimer()
        self.timer.setInterval(1)
        self.bind_slots()  # 绑定槽函数

    def image_pred(self, file_path): # 图片检测
        results = self.model(file_path) 
        image = results.render()[0]
        return convert2QImage(image)

    def open_image(self): # 打开图片
        print("点击了检测图片按钮")
        self.timer.stop()  # 停止视频检测
        file_path = QFileDialog.getOpenFileName(self, dir="./data02/images/train", filter="*.jpg;*.png;*.jpeg")
        if file_path[0]:
            file_path = file_path[0]
            qimage = self.image_pred(file_path)
            self.input.setPixmap(QPixmap(file_path))
            self.output.setPixmap(QPixmap.fromImage(qimage))

    def video_pred(self):  # 视频检测
        ret, frame = self.video.read()
        if not ret:
            self.timer.stop()
        else:
            frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
            self.input.setPixmap(QPixmap.fromImage(convert2QImage(frame)))
            results = self.model(frame)
            image = results.render()[0]
            self.output.setPixmap(QPixmap.fromImage(convert2QImage(image)))

    def open_video(self):  # 打开视频
        print("点击了检测视频！")
        file_path = QFileDialog.getOpenFileName(self, dir="./data02", filter="*.mp4")
        if file_path[0]:
            file_path = file_path[0]
            self.video = cv2.VideoCapture(file_path)
            self.timer.start()

    def bind_slots(self):  # 绑定槽函数
        self.det_image.clicked.connect(self.open_image)
        self.det_video.clicked.connect(self.open_video)
        self.timer.timeout.connect(self.video_pred)


if __name__ == "__main__":
    app = QApplication(sys.argv)

    window = MainWindow()
    window.show()

    app.exec()

Gradio builds Web GUI

InstallGradio

Activate the previous virtual environmentyolov5

Enter the following command in the terminal of this environment

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple gradio

accomplish

Create new gradio_demo.pyfile

Input picture, output picture, prediction function

Gradio does not support real-time detection

import torch
import gradio as gr

model = torch.hub.load("./", "custom", path="runs/train/exp/weights/best.pt", source="local")

title = "基于Gradio的YOLOv5演示项目"

desc = "这是一个基于Gradio的YOLOv5演示项目，非常简洁，非常方便！"

base_conf, base_iou = 0.25, 0.45


def det_image(img, conf_thres, iou_thres):
    model.conf = conf_thres
    model.iou = iou_thres
    return model(img).render()[0]


gr.Interface(
    inputs=["image", gr.Slider(minimum=0, maximum=1, value=base_conf), gr.Slider(minimum=0, maximum=1, value=base_iou)],
    outputs=["image"],
    fn=det_image,
    title=title,
    description=desc,
    live=True,
    examples=[["./data02/images/train/30.jpg", base_conf, base_iou],
              ["./data02/images/train/60.jpg", 0.3, base_iou]]
).launch(share=True)

If you want to publish to the Internet, set the share=True parameter, but when running, the following error will be reported due to network reasons. You can download the corresponding file through the prompted URL, rename it, put it in the corresponding directory, and run it again.