Min Y.-Z.,Lanzhou Jiaotong University |
Kang F.,Lanzhou Jiaotong University |
Dang J.-W.,Lanzhou Jiaotong University |
Zhang Z.-H.,Lanzhou Jiaotong University |
And 2 more authors.
Tiedao Xuebao/Journal of the China Railway Society | Year: 2014
Precise light spot center positioning technology is critical to measurement accuracy of the image ballastless track settlement monitoring system. With traditional light spot positioning technology, large errors may be produced when positioning disturbed irregular spot images on received target surfaces. To deal with the problem, the image was preprocessed, and on this basis, according to the characteristic that gradations of the acquired image appear in Gaussian distribution, the gray-scale distribution curve fitting method based on the gravity center of spot was put forward. With the proposed method, firstly, the light spot center was initially positioned by using the gravity center method; then the laser spot center was positioned precisely by using the quadratic curve to fit grayscale Gaussian distribution. Laboratory and field tests show that the proposed method is of high positioning accuracy and good robustness, and is appropriate to satisfy the accuracy requirements of image-type ballastless track settlement monitoring systems.