基于改进SSD模型的路面病害识别算法研究  

Research on pavement disease recognition algorithm based on improved SSD model

作　　者：黄凯枫 张博熠 王梦 刘庆华 HUANG Kaifeng;ZHANG Boyi;WANG Meng;LIU Qinghua(School of Computer Science,Jiangsu University of Science and Technology,Zhenjiang 212100,China)

机构地区：[1]江苏科技大学计算机学院,镇江212100

出　　处：《江苏科技大学学报（自然科学版）》2023年第2期53-60,共8页Journal of Jiangsu University of Science and Technology:Natural Science Edition

基　　金：国家自然科学基金资助项目(51008143);江苏省六大高峰人才项目(XYDXX-117)。

摘　　要：为解决公路路面病害图像特征不突出、检测精度偏低的问题,提出了一种改进SSD模型的路面病害识别算法.在SSD网络结构的基础上将其基础网络替换为Dense-net网络,使得特征信息更加容易被获取,并能够降低网络参数数量.同时在算法中增添了注意力机制,加强有用特征的利用效率.为了更好地观察算法改进的效果,不仅在已知的路面数据集上进行了测试,还在自制的数据集上进行了测试.从测试结果来看,SSD模型改进后在两种数据集上的分类准确度分别为93.5%和90.28%,比原SSD300模型分别提高了4.8%和6.36%,说明该模型能够有效的提升病害识别的准确性.In order to solve the problem that the image features of highway pavement diseases are not prominent and the detection accuracy is low,this paper proposes an improved SSD model of pavement disease recognition algorithm.The basic network of the SSD network is replaced with the Dense-net network on the basis of retaining the SSD network structure,which makes the characteristic information easier to obtain and can reduce the number of network parameters.At the same time,an attention mechanism is added to the algorithm to enhance the utilization efficiency of useful features.In order to better observe the effect of the algorithm improvement,we tested on not only the known road data set,but also the self-made data set.Judging from the test results,the improved SSD model has a classification accuracy of 93.5%and 90.28%on the two data sets,a significant improvement(4.8%and 6.36%,respectively)compared with the original SSD300 model.This model can effectively improve the accuracy of disease recognition.

关 键 词：路面病害 目标检测 神经网络 改进SSD 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]