The practicality of online calibration algorithms in actual autonomous driving scenarios is enhanced by proposing an online calibration method for intelligent networked automotive lidar and camera based on depth-edge matching. The initial values of external parameters are estimated and calculated through hand-eye calibration. The solution of hand-eye calibration is optimized and accurate external parameters are obtained through data conversion. The CMA-ES algorithm is utilized to optimize the optimized parameters which are further compared with the conventional method based on edge matching. It is found that the provided frames of data, the external parameters can be appropriately improved by the method in this paper, and the algorithm congregates in about 1000 seconds. However, the conventional method cannot optimize the parameters correctly when there are only 2 frames of data. The rotation error of most results of this method is between 0.1° and 0.8°, and the translation error is between 0.02m and 0.06m. Compared with other representative algorithms of various methods, the errors in all aspects are more balanced and there is no outstanding error value.

In the modern world, it is difficult to prevent terrorism due to the relatively closed environment, dense personnel, large passenger flow, long line and wide coverage of urban rail transit. Identity recognition is a core element of security. The design and study of an urban rail transit security system based on face recognition technology are proposed in this paper. Through the study on the face recognition algorithm of intelligent security systems in urban rail transit, the related introduction of face recognition technology is done. The analysis of the main mode of face recognition is carried out utilizing the practical application design ideas. The results by experimental analysis show that if FAR is set to a very low range (such as 0.1% or even 0.01%) meanwhile FRR can reach a very low level (such as less than 1%). Such a system has practical value and otherwise, it may face a large number of passenger affairs and complaints to be handled. When FAR is set to 0.1% and N is 1.6 million, FRR can reach 2.1%. However, according to the test, when the picture quality deteriorates (during image captured by a webcam), the FRR will increase by 2 to 3 times. If a Webcam is used for recognition in Mugshot, the lowest FRR of the three top algorithms is only 5.21%.Povzetek: Tehnologija prepoznavanja obrazov je uporabljena za nadzor osumljencev -teroristov na vlakih.