A method of target detection in optical remote sensing images based on Bayesian transfer learning

Abstract

The object detection method of optical remote sensing images based on Bayesian migration learning provided by the present invention uses the Fisher information matrix to represent the knowledge contained in the pre-trained target detector on the source data set, and uses the Fisher information matrix The information matrix constructs the objective function of the target detector. During the training process, this item will participate in the learning of the target detector on the target data set, so that it can retain the learned knowledge to a certain extent and improve the detection accuracy; Compared with other optical remote sensing image target detection algorithms, the present invention effectively improves the accuracy of optical remote sensing image target detection without introducing additional parameters to be learned; it can efficiently and accurately assist human interpretation and save manpower.

Description

technical field

[0001] The invention belongs to the technical field of image processing, and further relates to an optical remote sensing image target detection technology, in particular to an optical remote sensing image target detection method based on Bayesian transfer learning. Background technique

[0002] Optical remote sensing image target detection is to judge whether an optical remote sensing image contains interesting targets such as aircraft, vehicles, ports, etc., and successfully locate and identify them. As one of the basic tasks in the field of remote sensing image analysis, it has important application value in environmental monitoring, land use, urban planning, transportation, military and other fields. However, due to the influence of external factors such as different shooting angles, complex background components, illumination and shadow changes, it is a great challenge to carry out target detection tasks on optical remote sensing images. In order to ove...

Images