Objective For computer-aided Chinese medical diagnosis and aiming at the problem of insufficient segmentation,a novel multi-level method based on the multi-scale fusion residual neural network(MF2ResU-Net)model is pro...Objective For computer-aided Chinese medical diagnosis and aiming at the problem of insufficient segmentation,a novel multi-level method based on the multi-scale fusion residual neural network(MF2ResU-Net)model is proposed.Methods To obtain refined features of retinal blood vessels,three cascade connected UNet networks are employed.To deal with the problem of difference between the parts of encoder and decoder,in MF2ResU-Net,shortcut connections are used to combine the encoder and decoder layers in the blocks.To refine the feature of segmentation,atrous spatial pyramid pooling(ASPP)is embedded to achieve multi-scale features for the final segmentation networks.Results The MF2ResU-Net was superior to the existing methods on the criteria of sensitivity(Sen),specificity(Spe),accuracy(ACC),and area under curve(AUC),the values of which are 0.8013 and 0.8102,0.9842 and 0.9809,0.9700 and 0.9776,and 0.9797 and 0.9837,respectively for DRIVE and CHASE DB1.The results of experiments demonstrated the effectiveness and robustness of the model in the segmentation of complex curvature and small blood vessels.Conclusion Based on residual connections and multi-feature fusion,the proposed method can obtain accurate segmentation of retinal blood vessels by refining the segmentation features,which can provide another diagnosis method for computer-aided Chinese medical diagnosis.展开更多
As an important research topic in recent years,semantic segmentation has been widely applied to image understanding problems in various fields.With the successful application of deep learning methods in machine vision...As an important research topic in recent years,semantic segmentation has been widely applied to image understanding problems in various fields.With the successful application of deep learning methods in machine vision,the superior performance has been transferred to agricultural image processing by combining them with traditional methods.Semantic segmentation methods have revolutionized the development of agricultural automation and are commonly used for crop cover and type analysis,pest and disease identification,etc.We frst give a review of the recent advances in traditional and deep learning methods for semantic segmentation of agricultural images according to different segmentation principles.Then we introduce the traditional methods that can effectively utilize the original image information and the powerful performance of deep learningbased methods.Finally,we outline their applications in agricultural image segmentation.In our literature,we identify the challenges in agricultural image segmentation and summarize the innovative developments that address these challenges.The robustness of the existing segmentation methods for processing complex images still needs to be improved urgently,and their generalization abilities are also insufficient.In particular,the limited number of labeled samples is a roadblock to new developed deep learning methods for their training and evaluation.To this,segmentation methods that augment the dataset or incorporate multimodal information enable deep learning methods to further improve the segmentation capabilities.This review provides a reference for the application of image semantic segmentation in the field of agricultural informatization.展开更多
基金Key R&D Projects in Hebei Province(22370301D)Scientific Research Foundation of Hebei University for Distinguished Young Scholars(521100221081)Scientific Research Foundation of Colleges and Universities in Hebei Province(QN2022107)。
文摘Objective For computer-aided Chinese medical diagnosis and aiming at the problem of insufficient segmentation,a novel multi-level method based on the multi-scale fusion residual neural network(MF2ResU-Net)model is proposed.Methods To obtain refined features of retinal blood vessels,three cascade connected UNet networks are employed.To deal with the problem of difference between the parts of encoder and decoder,in MF2ResU-Net,shortcut connections are used to combine the encoder and decoder layers in the blocks.To refine the feature of segmentation,atrous spatial pyramid pooling(ASPP)is embedded to achieve multi-scale features for the final segmentation networks.Results The MF2ResU-Net was superior to the existing methods on the criteria of sensitivity(Sen),specificity(Spe),accuracy(ACC),and area under curve(AUC),the values of which are 0.8013 and 0.8102,0.9842 and 0.9809,0.9700 and 0.9776,and 0.9797 and 0.9837,respectively for DRIVE and CHASE DB1.The results of experiments demonstrated the effectiveness and robustness of the model in the segmentation of complex curvature and small blood vessels.Conclusion Based on residual connections and multi-feature fusion,the proposed method can obtain accurate segmentation of retinal blood vessels by refining the segmentation features,which can provide another diagnosis method for computer-aided Chinese medical diagnosis.
基金the Post-graduate's Innovation Fund Project of Hebei University(HBU2022ss037)the High-Performance Computing Center of Hebei University.
文摘As an important research topic in recent years,semantic segmentation has been widely applied to image understanding problems in various fields.With the successful application of deep learning methods in machine vision,the superior performance has been transferred to agricultural image processing by combining them with traditional methods.Semantic segmentation methods have revolutionized the development of agricultural automation and are commonly used for crop cover and type analysis,pest and disease identification,etc.We frst give a review of the recent advances in traditional and deep learning methods for semantic segmentation of agricultural images according to different segmentation principles.Then we introduce the traditional methods that can effectively utilize the original image information and the powerful performance of deep learningbased methods.Finally,we outline their applications in agricultural image segmentation.In our literature,we identify the challenges in agricultural image segmentation and summarize the innovative developments that address these challenges.The robustness of the existing segmentation methods for processing complex images still needs to be improved urgently,and their generalization abilities are also insufficient.In particular,the limited number of labeled samples is a roadblock to new developed deep learning methods for their training and evaluation.To this,segmentation methods that augment the dataset or incorporate multimodal information enable deep learning methods to further improve the segmentation capabilities.This review provides a reference for the application of image semantic segmentation in the field of agricultural informatization.
