Dynamic object sensing and predicting method based on multi-source sensor information fusion

Abstract

The invention discloses a dynamic object sensing and predicting method based on multi-source sensor information fusion, and the method comprises the following steps: detecting the type of a target object through a multi-source sensor, and obtaining color point cloud data; estimating the position of the target object through the multi-source sensor; and performing accurate internal reference and external reference calibration on the target object through a calibration method by a calibration module of the multi-source sensor; calculating the position of the target object in the three-dimensional coordinate system; enabling the multi-source sensor to attach a timestamp to the information in the above steps; obtaining the motion direction and speed of the target object through the time sequence change of the position of the target object; and according to the change of the position and the speed, predicting the collision possibility. Compared with the prior art, the invention has the following advantages: wide-area sensing information can be provided for an unmanned vehicle or an autonomous robot through the information such as the type, the position, the speed and the motion trail of the dynamic object in the scene obtained through sensing prediction, so that the safety and the efficiency of operation are improved.

Description

technical field [0001] The invention relates to the technical fields of line computer vision and intelligent transportation, in particular to a dynamic object perception and prediction method based on multi-source sensor information fusion. Background technique [0002] With the rapid development of cars as a means of transportation in daily life, automatic driving and assisted driving functions have become the development trend of new smart cars. However, recent frequent accidents caused by autonomous driving and assisted driving pose new challenges to the safety of unmanned or assisted driving. Limited by the blind area of ​​vision, in order to prevent unmanned vehicles from colliding with nearby obstacles while driving, existing solutions often use infrared laser solutions to ensure the safety of vehicles during driving through distance measurement. However, there are two problems in this scheme at present: (1) The limitation of infrared laser sensor: infrared laser rang...

AI Technical Summary

Problems solved by technology

However, there are two problems in this scheme at present: (1) The limitation of infrared laser sensor: infrared laser ranging can provide the distance information of the nearest obstacle more accurately, but for the category of the measured object, whether it is a dynamic object, Whether there is no further perception of potential threats to the current driver

(2) Limitations of a single sensor: a single infrared sensor can only measure distance information, and the information it can provide is limited, while cameras are often greatly affected by field of view and climate and weather changes. The temperature information provided helps to determine whether the nearby dynamic object is a vehicle or a pedestrian with a heat source, but it has not been applied to the perception of vehicles so far.

Images