yolov8+tensorrt+pyqt5+多线程+鼠标单机放大缩小多路视频监控系统+账户密码验证登录界面（持续开发完善中）

一.最近公司做多路检测项目，b站找了好多类似功能实现，源码都收费，好贵，无奈，自己搞

    1.废话不多说，先看demo吧

                

                  

                

 

        我这里算法用的是官方的yolov8s哈，各位大佬可以结合自己的算法转成engine再部署，另外我这里用的转化工具是参考github：

             https://github.com/triple-Mu/YOLOv8-TensorRT 

        另外我的多线程中的推理部分也用的这个大神中间的部分代码，自己做了一些修改，特此鸣谢。

        另注：我这个还没写完哈，我只有一路是摄像头，其他都是视频文件模拟的，我的是显卡是3080，i5的12代cpu，后面还要加的功能很多，因为落地摄像头是rtsp流，可能还要加cuda硬编解码做优化。基本功能已完成，后面会持续优化而且代码也会有优化，现在只相当于草稿了，需要的留意更新吧。。。。。

v1.1.1

pyqt5 账户密码登录界面

class loginUi(QMainWindow,login_ui):

    def __init__(self):

        QMainWindow.__init__(self)
        self.setupUi(self)
        self.pushButton.clicked.connect(self.checkLogin)
        self.pushButton_2.clicked.connect(self.exitSys)

    def exitSys(self):
        sys.exit()

    def checkLogin(self):

        self.name = self.lineEdit_2.text()
        self.passwd = self.lineEdit_3.text()

        self.getSql()
        if self.name == self.nameSave and self.passwdSave == self.passwd:
            self.myPlayer = mainUi()
            self.myPlayer.show()
            self.close()
        else:
            self.errorUi = errorLogin()
            self.errorUi.show()
            self.close()

    def getSql(self):
        with open("lo.txt",'r') as f:
            lines = f.readlines()
            self.nameSave = lines[0][:-1]
            self.passwdSave = lines[1]
            f.close()

class errorLogin(QMainWindow,error_ui):

    def __init__(self):

        QMainWindow.__init__(self)
        self.setupUi(self)
        self.pushButton.clicked.connect(self.returnUi)

    def returnUi(self):
        self.log = loginUi()
        self.log.show()
        self.close()

v1.0.0

class mainUi(QMainWindow,convert_ui):

    def __init__(self):

        QMainWindow.__init__(self)
        self.setupUi(self)
        self.pushButton.clicked.connect(self.openVideo)
        self.pushButton_2.clicked.connect(lambda :self.changeLabel(1))
        self.pushButton_3.clicked.connect(lambda: self.changeLabel(2))
        self.pushButton_4.clicked.connect(lambda: self.changeLabel(3))
        self.pushButton_5.clicked.connect(lambda: self.changeLabel(4))
        self.pushButton_6.clicked.connect(lambda: self.changeLabel(5))
        self.pushButton_7.clicked.connect(lambda: self.changeLabel(6))
        self.pushButton_8.clicked.connect(lambda: self.changeLabel(7))
        self.pushButton_9.clicked.connect(lambda: self.changeLabel(8))
        self.pushButton_10.clicked.connect(self.exitSys)

        self.countButton1 = 0
        self.countButton2 = 0
        self.countButton3 = 0
        self.countButton4 = 0
        self.countButton5 = 0
        self.countButton6 = 0
        self.countButton7 = 0
        self.countButton8 = 0

    def exitSys(self):

        sys.exit()

    def getArr(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0],QImage.Format_BGR888)
        self.label.setPixmap(QPixmap(img))
        self.label.setScaledContents(True)

    def getArr2(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_2.setPixmap(QPixmap(img))
        self.label_2.setScaledContents(True)

    def getArr3(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_3.setPixmap(QPixmap(img))
        self.label_3.setScaledContents(True)

    def getArr4(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_4.setPixmap(QPixmap(img))
        self.label_4.setScaledContents(True)

    def getArr5(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_5.setPixmap(QPixmap(img))
        self.label_5.setScaledContents(True)

    def getArr6(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_6.setPixmap(QPixmap(img))
        self.label_6.setScaledContents(True)

    def getArr7(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_7.setPixmap(QPixmap(img))
        self.label_7.setScaledContents(True)

    def getArr8(self,arr):
        img = QImage(arr, arr.shape[1], arr.shape[0], QImage.Format_BGR888)
        self.label_8.setPixmap(QPixmap(img))
        self.label_8.setScaledContents(True)

    def openVideo(self):

        self.pushButton.setEnabled(False)
        self.pushButton.setText("请等待")


        self.imgGet = inferVideo1()
        self.imgGet.updateImg.connect(self.getArr)
        self.imgGet.start()

        self.imgGet2 = inferVideo2()
        self.imgGet2.updateImg1.connect(self.getArr2)
        self.imgGet2.start()

        self.imgGet3 = inferVideo3()
        self.imgGet3.updateImg2.connect(self.getArr3)
        self.imgGet3.start()

        self.imgGet4 = inferVideo4()
        self.imgGet4.updateImg3.connect(self.getArr4)
        self.imgGet4.start()

        self.imgGet5 = inferVideo5()
        self.imgGet5.updateImg4.connect(self.getArr5)
        self.imgGet5.start()

        self.imgGet6 = inferVideo6()
        self.imgGet6.updateImg5.connect(self.getArr6)
        self.imgGet6.start()

        self.imgGet7 = inferVideo7()
        self.imgGet7.updateImg6.connect(self.getArr7)
        self.imgGet7.start()

        self.imgGet8 = inferVideo8()
        self.imgGet8.updateImg7.connect(self.getArr8)
        self.imgGet8.start()

        self.pushButton.setStyleSheet("border-color : yellow")
        self.pushButton.setText("已播放")
        def returnPos(self):

        self.label.resize(501, 221)
        self.label.raise_()

        self.label_2.move(501, 0)
        self.label_2.resize(501, 221)
        self.label_2.raise_()

        self.label_3.move(0, 221)
        self.label_3.resize(501, 221)
        self.label_3.raise_()

        self.label_4.move(501, 221)
        self.label_4.resize(501, 221)
        self.label_4.raise_()

        self.label_5.move(0, 442)
        self.label_5.resize(501, 221)
        self.label_5.raise_()

        self.label_6.move(501, 442)
        self.label_6.resize(501, 221)
        self.label_6.raise_()

        self.label_7.move(0, 663)
        self.label_7.resize(501, 221)
        self.label_7.raise_()

        self.label_8.move(501, 663)
        self.label_8.resize(501, 221)
        self.label_8.raise_()

    def changeLabel(self,x):

        if x == 1:

            self.countButton1 += 1

            if self.countButton1 % 2 == 0:

                self.returnPos()

            else:

                self.label.resize(self.label.width() + 501, self.label.height() + 663)
                self.label.raise_()

        elif x == 2:

            self.countButton2 += 1

            if self.countButton2 % 2 == 0:

                self.returnPos()

            else:

                self.label_2.move(0, 0)
                self.label_2.resize(self.label_2.width() + 501, self.label_2.height() + 663)
                self.label_2.raise_()

        elif x == 3:

            self.countButton3 += 1

            if self.countButton3 % 2 == 0:

                self.returnPos()

            else:

                self.label_3.move(0, 0)
                self.label_3.resize(self.label_3.width() + 501, self.label_3.height() + 663)
                self.label_3.raise_()

        elif x == 4:

            self.countButton4 += 1

            if self.countButton4 % 2 == 0:

                self.returnPos()

            else:

                self.label_4.move(0, 0)
                self.label_4.resize(self.label_4.width() + 501, self.label_4.height() + 663)
                self.label_4.raise_()

        elif x == 5:

            self.countButton5 += 1

            if self.countButton5 % 2 == 0:

                self.returnPos()

            else:

                self.label_5.move(0, 0)
                self.label_5.resize(self.label_5.width() + 501, self.label_5.height() + 663)
                self.label_5.raise_()

        elif x == 6:

            self.countButton6 += 1

            if self.countButton6 % 2 == 0:

                self.returnPos()

            else:

                self.label_6.move(0, 0)
                self.label_6.resize(self.label_6.width() + 501, self.label_6.height() + 663)
                self.label_6.raise_()

        elif x == 7:

            self.countButton7 += 1

            if self.countButton7 % 2 == 0:

                self.returnPos()

            else:

                self.label_7.move(0, 0)
                self.label_7.resize(self.label_7.width() + 501, self.label_7.height() + 663)
                self.label_7.raise_()

        elif x == 8:

            self.countButton8 += 1

            if self.countButton8 % 2 == 0:

                self.returnPos()

            else:

                self.label_8.move(0, 0)
                self.label_8.resize(self.label_8.width() + 501, self.label_8.height() + 663)
                self.label_8.raise_()

   

def det_postprocess(data: Tuple[Tensor, Tensor, Tensor, Tensor]):

    assert len(data) == 4
    num_dets, bboxes, scores, labels = data[0][0], data[1][0], data[2][
        0], data[3][0]
    nums = num_dets.item()
    if nums == 0:
        return bboxes.new_zeros((0, 4)), scores.new_zeros((0, )), labels.new_zeros((0, ))
    bboxes = bboxes[:nums]
    scores = scores[:nums]
    labels = labels[:nums]

    return bboxes, scores, labels

def capRun(Img,W,H,device,Engine):
    draw = Img.copy()
    bgr, ratio, dwdh = letterbox(Img, (W, H))
    tensor = blob(bgr, return_seg=False)
    dwdh = torch.asarray(dwdh * 2, dtype=torch.float32, device=device)
    tensor = torch.asarray(tensor, device=device)
    data = Engine(tensor)
    bboxes, scores, labels = det_postprocess(data)
    if bboxes.numel() == 0:
        # if no bounding box
        return draw
    bboxes -= dwdh
    bboxes /= ratio
    for (bbox, score, label) in zip(bboxes, scores, labels):
        bbox = bbox.round().int().tolist()
        cls_id = int(label)
        cls = CLASSES[cls_id]
        color = COLORS[cls]
        cv2.rectangle(draw, bbox[:2], bbox[2:], color, 2)
        cv2.putText(draw, f'{cls}:{score:.3f}', (bbox[0], bbox[1] - 2), cv2.FONT_HERSHEY_SIMPLEX,
                    0.75, [225, 255, 255], thickness=2)
    return draw

class inferVideo1(QThread):

    updateImg = pyqtSignal(np.ndarray)

    def __init__(self):

        super(inferVideo1,self).__init__()
        self.device = torch.device("cuda:0")
        self.Engine = TRTModule("yolov8s.engine", self.device)
        self.H, self.W = self.Engine.inp_info[0].shape[-2:]
        self.Engine.set_desired(['num_dets', 'bboxes', 'scores', 'labels'])

    def run(self):

        cap = cv2.VideoCapture(0)

        if not cap.isOpened():

            print("Unable to open camera")

        else:

            while True:

                ret, Img = cap.read()

                if ret:

                    draw = capRun(Img,self.W,self.H,self.device,self.Engine)
                    self.updateImg.emit(draw)

                else:

                    break

        cap.release()

上面是一个线程的代码哈，需要的自己复制粘贴8个就行。我的i5的12代 16g内存 完美运行。。。

需要源码的可以私信我哈。