Three-dimensional target detection method and device and storage medium

Abstract

The invention provides a three-dimensional target detection method and device and a storage medium, and the method comprises the steps: setting a first coordinate center of a target object in a monocular image as a second coordinate center of a 3D bounding box; setting space coordinate constraints of the 3D bounding box according to external parameters and internal parameters, setting a directionloss function and a size loss function of the 3D bounding box, and generating a model loss function; and training the convolutional neural network model by using the monocular image training sample and based on the spatial coordinate constraint and the model loss function so as to perform three-dimensional target detection processing on the monocular image. The spatial coordinate constraint, the direction loss function and the size loss function are set, and the convolutional neural network model is trained to construct the multi-task neural network, so that 3D target detection of a monocularimage can be realized; the efficiency and precision of three-dimensional target detection can be improved, and the use cost is reduced.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a three-dimensional object detection method, device and storage medium. Background technique [0002] Target detection is a traditional task in the field of computer vision. Different from image recognition, target detection not only needs to identify the objects existing on the image and give the corresponding category, but also needs to pass the position of the object through the minimum bounding box (Bounding box) way give. At present, most 3D object detection methods rely on lidar data to provide accurate depth information in autonomous driving scenarios, but lidar is expensive and has a relatively small field of view; monocular cameras are complex to install and low in cost, but currently There is no technical solution for 3D object detection based on monocular vision. Contents of the invention [0003] In view of this, a technical problem to be solved b...

AI Technical Summary

Problems solved by technology

At present, most 3D object detection methods rely on lidar data to provide accurate depth information in autonomous driving scenarios, but lidar is expensive and has a relatively small field of view; monocular cameras are complex to install and low in cost, but currently There is no 3D target detection technology solution based on monocular vision

Images