Optimization Of Embedded Point Switch Detection System

Currently in the turnout gap monitoring system used in the field of railway,the camera lens for catching gap image will have some degree of radial distortion during the production process.The design and installation of the camera capture device as well as the impact of the traffic vibrations may cause that the camera lens are not perpendicular to the shooting plane.So the gap image will generate trapezoidal distortion.All of these will not only have an effect on the accuracy of the whole system,but also lead to the failure of detecting the gap images in severe case.The camera capture device is installed in the closed internal switch machine.The light source is entirely provided by the LED flash on the tiny camera when capturing images.It is prone to uneven exposure.Besides,the structures inside the switch machine are made of smooth metal,which has a strong light reflection and diffuse reflection.So the gap images will have a strong interference effect,and the procession of the system will be decreased.The monitoring host machine is not closely associated with the field capture devices in the existing turnout gap monitoring system,so it can't judge the operating status of the field capture devices.When they run into the wrong condition,it needs the maintenance person to inspect and recover.This not only increases the burden for field maintenance work,but also affects the real-time operation of the railway.This paper studies the characteristics of the existing turnout gap monitoring system seriously and designs the reliability of the system integrating varieties of technologies and equipment including digital image processing technology,embedded equipment and heartbeat interaction mechanism,reliability of the machine-version-based embedded turnout gap monitoring system is optimized.This paper introduces the linear and nonlinear camera model and summarizes the research status of image distortion correction.Combined with the characteristics of the existing turnout gap monitoring system,it finishes the correction of the radial distortion using an arc surface mapping method without camera calibration.This paper also makes a correction for' the trapezoidal distortion and realizes the identification by resetting the detecting area.They both improve the accuracy of the system.This paper designs two sets of image recognition algorithm to eliminate the light pollution:ZD6 type and S700K type,then makes a detailed description of their principle and process.So the gap image polluted by light can be effectively identified.Finally,this paper finishes the design of the heartbeat interaction mechanism based on the sending control commands of the turnout gap monitoring system.It not only ensures that the operating state of the monitoring device is known by the monitoring machine at any time,but also achieves a soft restart of the monitoring device from remote place.So the maintenance people don't have to go to the field area to restart the monitoring device.