检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:罗佳辉 方承志[1] 杨豪 LUO Jiahui;FANG Chengzhi;YANG Hao(School of Electronic and Optical Engineering,School of Flexible Electronics(Future Technology),Nanjing University of Posts and Telecommunications,Nanjing 210023)
机构地区:[1]南京邮电大学电子与光学工程学院、柔性电子(未来技术)学院,南京2210023
出 处:《计算机与数字工程》2024年第12期3728-3734,共7页Computer & Digital Engineering
基 金:国家自然科学基金项目(编号:61977039)资助。
摘 要:为准确区分自然场景中的相邻文本实例,快速精确定位文字实例,提出一种非局部特征增强的文本检测算法。该算法在DBNet的基础上,以轻量级网络resnet-18作为骨干网络,采用特征金字塔增强模块(FPEM)和特征金字塔融合模块(FFM)来弥补轻量级网络提取特征能力的不足,针对传统网络处理区域的局限性问题,加入改进的非局部网络(GCNet),使模型通过全局的角度来捕获图像信息。采用可微二值化优化分割网络来区分紧密的文本实例,同时简化了后处理。针对数据样本不均衡的问题,使用focal loss作为损失函数,调整正负样本权重并使模型更加关注困难样本。实验结果表明,在ICDAR2015数据集上,该算法比目前先进的DBNet在F值和检测速度上均有一定的提升。To precisely distinguish adjacent text instances in natural scenes,quickly and accurately locate text instances,a text detection algorithm with non-local feature enhancement is proposed.Based on DBNet,the algorithm takes the lightweight net-work resnet-18 as the backbone network,and adopts feature pyramid enhancement module(FPEM)and feature pyramid fusion module(FFM)to compensate the insufficiency of feature extraction capabilities of lightweight networks.Considering the limitation of traditional network processing area,an improved non global context net(GCNet)is introduced,so as to facilitate the model to capture image information from a global perspective.Meanwhile,a differentiable binary optimization segmentation network is adopt-ed to distinguish close text instances,and the post-processing is simultaneously simplified.For the unbalanced data samples,the fo-cal loss is selected as the loss function,the weights of positive and negative samples are adjusted,and pay more attention to difficult samples.The experiment results show that the algorithm achieves a certain improvement on value F and detection speed on IC-DAR2015 data set if comparing with current advanced DBNet.
关 键 词:文本检测 特征增强 改进的非局部网络 可微二值化 focal loss
分 类 号:TP391[自动化与计算机技术—计算机应用技术]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.249