Vehicle target joint cognition method and system based on point cloud and image data

Abstract

The invention provides a vehicle target joint cognition method and system based on point cloud and image data. The system comprises a data level combination module, a deep learning target detection wood block and a combination cognition module. The data level combination module obtains three-dimensional point cloud data and image data, and fuses the point cloud data and the image data. Fusion datais collected by the deep learning target detection module to be subjected to feature level detection and recognition, a detection result is output. The joint cognition module judges the feature levelfusion detection result and the data level fusion detection result by adopting an evidence theory method, reliability distribution is obtained to serve as output, and detection stability and robustness are improved.

Description

technical field [0001] The invention relates to the field of unmanned driving technology, in particular to a method and system for joint recognition of vehicle targets based on point cloud and image data. Background technique [0002] An unmanned vehicle is a vehicle with autonomous driving behavior. On the basis of traditional vehicles, artificial intelligence modules such as environmental perception, intelligent decision-making, path planning, and behavior control are added, and then it can interact with the surrounding environment and rent out corresponding functions. Decision-making and action of mobile wheeled robots. [0003] Thanks to the rapid development of new sensor technology and its learning technology, a variety of sensors are used in unmanned driving to perform complete, accurate, robust and real-time perception of the surrounding environment. The process of environment perception mainly includes sensor calibration, structured road detection, unstructured roa...

AI Technical Summary

Problems solved by technology

[0004] The existing technology can combine the detection of optical images and three-dimensional point cloud data, and match the image features and three-dimensional point cloud features through the convolutional neural network, so as to obtain the outer rectangular frame of the target, which is used for the detection of the target position. However, the existing methods are only suitable for the recognition and detection of car targets, not suitable for pedestrians and bicycles, or do not fully combine 3D point cloud data, so the accuracy is low and cannot be directly applied to unmanned driving

Images