Semantic segmentation of very high-resolution remote sensing image based on multiple band combinations and patchwise scene analysis

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

et al. 2021

Self Cite

The classification of remote sensing images with high spatial resolution requires considerable training samples, but the process of sample making is slow and laborious. How to guarantee the accuracy of supervised classification under the condition of limited samples is an urgent problem to be solved in the field of supervised classification. For addressing this problem, we proposes an improved conditional generative adversarial network with edge feature (ECGAN) to augment limited training data for the classification of remote sensing images with high spatial resolution in this paper. On the basis of conditional generative adversarial network, feature factors of interclass boundaries and intraclass edges are added to networks, and an objective function with multiscale and multilevel features is constructed. The ISPRS Potsdam and Vaihingen remote sensing datasets are regarded as examples. Results indicate that the high-resolution remote sensing images generated by using the network proposed in this paper have abundant texture, accurate edges, and are highly similar to real images. The generated images are used to augment training samples, and an experiment for classifying high-resolution remote sensing images is conducted. The classification results of the proposed augmentation method perform better than that of the traditional sample augmentation method. We prove that ECGAN as a means of sample augmentation can effectively solves the problem that the classification effect is unideal when the supervised classification sample is insufficient.

show abstract

Section: B Evaluation Methods and Indicator 1) Evaluation Methodsmentioning

confidence: 99%

ECGAN: An Improved Conditional Generative Adversarial Network With Edge Detection to Augment Limited Training Data for the Classification of Remote Sensing Images With High Spatial Resolution

Sui

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

et al. 2021

Self Cite

show abstract

“…Subsequently, an ASPP including 3×3 atrous convolutions with rate=2, 3, 4 and 1×1 conventional convolution is used in this encoder structure to handle multi-scale semantic information. The ASPP structure has an excellent classification effect on high spatial resolution remote sensing images [45,51,52] by using a convolutional feature layer with filters at multiple sampling rates to capture context information. In addition, the image pooling feature in ASPP is the mean value of the output of the previous layer.…”

Section: B 3d-resnet Encoder-decoder Architecturementioning

confidence: 99%

“…Such an algorithm can effectively reduce or even eliminate noise and provide new possibilities for VHR-HSI image segmentation. At present, many excellent semantic segmentation algorithms have been successfully applied in ordinary natural digital photos [38][39][40][41][42] and remote sensing images [43][44][45]. On this basis, researches use the combination of geographic object-based image analysis [46] or super-pixel [47] with a semantic segmentation algorithm and achieve good classification results for very high-resolution remote sensing images.…”

Section: Introductionmentioning

confidence: 99%

An Effective Classification Method for Hyperspectral Image With Very High Resolution Based on Encoder–Decoder Architecture

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Liu

et al. 2021

Self Cite

“…On the basis of image empty spectrum and texture information, contextual information is fully considered, showing a strong classification ability. Currently, the excellent semantic segmentation algorithms include FCN [22], PSPNet [23], Segnet [24], and DeepLab series [25][26][27][28]. This study designed a deep network model on the basis of DeepLab V3 that can identify offshore farming in the east coastal zone of Shandong Province, China.…”

Section: Introductionmentioning

confidence: 99%

A Modeling Method for Automatic Extraction of Offshore Aquaculture Zones Based on Semantic Segmentation

Sui

et al. 2020

IJGI

Self Cite