Small target detecting method based on R-FCN

Abstract

The invention discloses a small target detecting method based on R-FCN, wherein the method relates to the field of image processing. The method comprises the steps of introducing a to-be-detected image into a convolutional network, successively performing characteristic extraction on a to-be-detected image through M network layers according to a sequence from a topmost layer of M network layers to a downmost layer and according to a sequence from the downmost layer of the M network layers to the topmost layer, generating characteristic mapping graphs with different scales, selecting an N characteristic mapping graphs into an RPN for performing foreground-and-background classification, determining the coordinate of a foreground area, processing a characteristic mapping block which corresponds with the coordinate of the foreground area for obtaining a characteristic vector; inputting each characteristic vector into a classifier for performing secondary classification, detecting whether the kind to which the characteristic vector is affiliated corresponds with a to-be-detected small target and outputting a detecting result. According to the small target detecting method, a manner of combining a top-down characteristic pyramid and a down-top characteristic pyramid is utilized for performing small target detection on the characteristic mapping graphs with different scales, thereby reducing report omission for the small target and improving detecting precision.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an R-FCN-based small target detection method. Background technique [0002] Pedestrian detection has a wide range of applications in areas such as intelligent assisted driving, intelligent monitoring, pedestrian analysis, and intelligent robots. Therefore, how to optimize the performance and speed of pedestrian detection has become an important topic. The more commonly used method of pedestrian detection is to detect pedestrians in the image to be tested by collecting image information such as infrared information and hyperspectral information of the image to be tested. This method does not directly use the RGB image of the image to be tested. Information requires specific instruments to collect infrared information and hyperspectral information, so the detection is not direct enough, the accuracy is not high, and the detection speed is not fast. With the development of deep conv...

AI Technical Summary

Problems solved by technology

[0003] The deep convolutional neural network continuously learns and adjusts on the basis of a large number of training samples, which makes it perform well in target detection tasks, but it is mainly used for the detection of some large-scale general targets. When used for pedestrian detection, since the target of pedestrians is small, it is prone to false negatives when searching on a single scale, and the detection accuracy is still not high enough

Images