雾天条件下智能勘探车环境感知方法的研究

Research on Environment Perception Method of Intelligent Exploration Vehicle under Foggy Conditions

在数字图像处理应用领域,许多原因会使图像质量在成像过程中退化,不利于图像的目标检测。为此,本文设计了智能勘探车系统,以解决雾天环境下智能勘探小车采集图像模糊不清的问题。该系统以STM32单片机为控制芯片,驱动ESP32-cam模块实时传输图像,再通过TCP/IP协议将图像传输至计算机。计算机端利用OpenCV实现暗通道先验图像去雾算法,对雾天条件下实时采集的图像进行增强处理。操作人员可以在电脑端实时观察勘探车采集的图像及其对应的去雾后图像。通过观察对比去雾前后的图像,图像中的目标特征信息会更加明显。操作人员可以在手机端利用蓝牙调试助手远程控制智能勘探车和云台,实现180°垂直和水平视野中心变化,使勘探车具备灵活的视野变化能力。智能勘探车配备照明模块,方便车辆夜间图像采集。

In digital image processing applicati ons,many reasons can cause degradation of image quality during imaging,which is not conducive to image target detection.This paper designs the intelligent exploration cart system and solve the problem of blurred images captured by the intelligent exploration cart in foggy environment.The intelligent exploration cart system uses STM32 microcontroller as the control chip to drive the ESP32-cam module to transmit images in real time,and then transmits the images to the computer through TCP/IP protocol.The computer side uses OpenCV to implement the dark channel a priori image defogging algorithm to enhance the images collected in real time under foggy conditions.The operator can observe the images acquired by the exploration vehicle and their corresponding post-defog images in real time at the computer side.By observing and comparing the before and after defogging images,the target feature information in the images will be more obvious.The operator can use Bluetooth debugging assistant to remotely control the intelligent exploration vehicle and the gimbal on the cell phone side to achieve 180°vertical and horizontal field of view center change,so that the exploration vehicle has the ability to change the field of view flexibly.The intelligent exploration vehicle is equipped with lighting module to facilitate vehicle image acquisition at night.