A Few-Sample UAV Image Recognition Method Based on Virtual Sample Generation

Abstract

The invention belongs to the field of machine learning, and in particular relates to a method for identifying a few-sample unmanned aerial vehicle image based on virtual sample generation with high practicability. The present invention shoots a short video of a drone flying with N frames from a long distance from the ground to the air by a camera device, obtains N drone areas as positive samples, and collects trees, buildings, clouds and birds in combination with other related videos. , kites, balloons and other small interference areas are used as negative samples as training sample sets. The beneficial effects of the present invention are: the effective information and diversity of the samples are increased without excessive distortion, thereby improving the generalization ability of the model; in the identification algorithm part, the fast DPM model adopts the feature of the fixed root filter input The location of graphs and anchor points, as well as the component model that conforms to the modularity of the UAV, improves operation speed and accuracy.

Description

technical field [0001] The invention belongs to the field of machine learning, and in particular relates to a method for identifying a few-sample unmanned aerial vehicle image based on virtual sample generation with high practicability. Background technique [0002] In recent years, the application and prospects of the drone industry have attracted the attention of various industries. The number of drones has exploded, and safety hazards have also followed. UAVs fly in disorder, violate citizens' privacy, are used for illegal and criminal activities, and threaten the safety of airways. It is an effective supervision method to use the ubiquitous ultra-clear cameras to identify and monitor UAVs, so it is of great practical significance to study image-based UAV detection and identification methods. On the one hand, the difficulty of UAV image recognition is that the image size is extremely small and the intra-class spacing is large. On the other hand, due to the lack of learn...

AI Technical Summary

Problems solved by technology

On the one hand, the difficulty of UAV image recognition lies in the extremely small image size and large intra-class spacing. The reasons are the extremely small UAV image size and large scale differences caused by long distances and large motion ranges, as well as the weather environment and flight altitude. and shooting angles also make the differences between UAV images obvious; on the other hand, due to the lack of learning samples for UAV recognition tasks, various types of UAVs, various scenes, and difficulty in obtaining samples, the sample set is not enough Sufficient, resulting in low generalization ability of the recognition model

Images