Target detection method based on a dense connection characteristic pyramid network

Abstract

The invention discloses a target detection method based on a dense connection characteristic pyramid network, and relates to the image processing and computer vision technology. The method comprises the steps of collecting an image data set labeled with a target bounding box and category information; constructing a dense connection feature pyramid network containing a Squeeze-Exciton structure capable of learning a dependency relationship between feature channels as a feature extraction backbone network; alternately training the RPN subnet and the R-FCN subnet to obtain a target detection model; and detecting a specific target in the image by using the model. The squeeze-Exciton structure and a dense connection structure are introduced into a feature extraction trunk network, characterization capability of the model is enhanced, adaptability of the model to targets of different sizes is enhanced through the feature pyramid structure, calculation sharing of the whole network model is achieved to the maximum degree through the R-FCN detection head, calculation resources are saved, and performance of the whole target detection model is improved. .

Description

technical field [0001] The invention relates to image processing and computer vision technology, in particular to a target detection method based on a densely connected feature pyramid network. Background technique [0002] The problem to be solved by vision can be summed up as "What is Where", that is, "what is where". Traditional image classification mainly solves the "What" problem, that is, to judge what kind of objects an image contains. But with the development of technology and the growth of application requirements, we not only pay attention to the simple classification of images, but also hope to accurately obtain the objects of interest and their positions in the images (generally given by the target bounding box), which is exactly the goal Detect the problem to be solved. Object detection has great practical value and application prospects. At the same time, object detection is also the cornerstone of computer vision research and the basis for solving other high...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problems of insignificant appearance differences and large size differences that may exist simultaneously during target detection

Images