基于不规则尺度区域光谱信息的高光谱图像亚像元定位.pdf

第 42 卷第 4 期2023 年 8 月红 外 与 毫 米 波 学 报J.Infrared Millim.WavesVol.42，No.4August，2023文章编号：1001-9014（2023）04-0538-08DOI：10.11972/j.issn.1001-9014.2023.04.016Sub-pixel mapping based on spectral information of irregular scale areas for hyperspectral imagesWANG Peng1，2，3，CHEN Yong-Kang3，ZHANG Gong3，WANG Hong-Ying4，ZHAO Chun-Lei5，HAN Ling6*（1.Key Laboratory of Southeast Coast Marine Information Intelligent Perception and Application，Ministry of Natural Resources，Zhangzhou Institute of Surveying and Mapping，Zhangzhou 363000，China；2.Anhui Province Key Laboratory of Physical Geographic Environment，Chuzhou University，Chuzhou 239000，China；3.College of Electronic and Information Engineering，Nanjing University of Aeronautics and Astronautics，Nanjing 210016，China；4.School of Management，Nanjing University of Posts and Telecommunications，Nanjing 210003，China；5.Key Laboratory of Meteorology and Ecological Environment of Hebei Province，Meteorological Institute of Hebei，Shijiazhuang 050021，China；6.Xi an Key Laboratory of Territorial Spatial Information，Changan University，Xi an 710064，China）Abstract：Sub-pixel mapping technology can analyze mixed pixels and realize the transformation from fractional images to fine a land-cover mapping image at the sub-pixel level.However，the spectral information used by the traditional sub-pixel mapping methods is usually constructed in a specified rectangular local window，and the spectral information of all bands is rarely used，affecting the performance of sub-pixel mapping.To solve this issue，sub-pixel mapping based on spectral information of irregular scale areas（SIISA）for hyperspectral images is proposed in this paper.The experimental results on three remote sensing images show the proposed SIISA outperforms the existing sub-pixel mapping methods.Key words：hyperspectral images，sub-pixel mapping，super-resolution mapping，spatial-spectral information，irregular scale areas基于不规则尺度区域光谱信息的高光谱图像亚像元定位王鹏1，2，3，陈永康3，张弓3，王弘颖4，赵春雷5，韩玲6*（1.漳州测绘学院 自然资源部东南沿海海洋信息智能感知与应用重点实验室，福建 漳州363000；2.滁州学院 实景地理环境安徽省重点实验室，安徽 滁州239000；3.南京航空航天大学 电子信息工程学院，江苏 南京210016；4.南京邮电大学 管理学院，江苏 南京210003；5.河北省气象科学研究所 河北省气象与生态环境重点实验室，河北 石家庄050021；6.长安大学 西安市国土空间信息重点实验室，陕西 西安710064）摘要：亚像元定位技术可以分析混合像元，并实现从丰度图像到亚像元级精细土地覆盖定位图像的转换。然而，传统的亚像元定位方法所使用的光谱信息通常在指定的矩形局部窗口中构造，并且很少使用所有波段的光谱信息，影响了亚像元定位的性能。为了解决这一问题，本文提出了一种基于不规则尺度区域光谱信息的高光谱图像亚像元定位方法（SIISA）。在三幅遥感图像上的实验结果表明，所提出的SIISA优于现有的亚像元定位方法。Received date：2022 07 12，revised date：2023 04 15 收稿日期：2022 07 12，修回日期：2023 04 15Foundation items：Supported by the Foundation of Anhui Province Key Laboratory of Physical Geographic Environment（2022PGE010）；The Fundamental Research Funds for the Central Universities，CHD（300102353508）；the Key Laboratory of Southeast Coast Marine Information Intelligent Perception and Application，MNR（22101）；National Natural Science Foundation of China（61801211）；Natural Science Foundation of Jiangsu Province（BK20221478）；Hong Kong Scholars Program（XJ2022043）；S&T Program of Hebei（21567624H）；Open Project Program of Key Laboratory of Meteorology and Ecological Environment of Hebei Province（Z202102YH）Biography：HANG Ling（1964），Female，Liaoning，Professor，Doctor.Research area involves remote sensing information processing*Corresponding author：E-mail：4 期 WANG Peng et al：Sub-pixel mapping based on spectral information of irregular scale areas for hyperspectral images关键词：高光谱图像；亚像元定位；超分辨制图；空间-光谱信息；不规则区域中图分类号：TP751 文献标识码：AIntroductionDue to its rich spectral information from hundreds of bands，hyperspectral images not only have been actively investigated by remote sensing scholars in recent years，but also widely utilized in many fields，such as burned-area mapping，flood inundation mapping，and forest cover monitor1.However，with the continuous improvement of spectral resolution of hyperspectral images，its spatial resolution will be affected，resulting in many mixed pixels2，3.Spectral unmixing could handle with these mixed pixels to obtain the abundance images including the proportional values of sub-pixels belonging to land-cover classes，but the specific spatial distribution information on land-cover classes still cannot be extracted4.To solve this issue，sub-pixel mapping which is as the subsequent processing technology of spectral unmixing is proposed.In sub-pixel mapping，each pixel is divided into S2 sub-pixels according to factor scale S，land-cover class labels are then assigned to sub-pixels to obtain the fine land cover-class mapping images at sub-pixel scale5.According to the method of obtaining the sub-pixel mapping results，there are two main types，the initialization-then-optimization sub-pixel mapping and soft-then-hard sub-pixel mapping.In the initialization-then-optimization sub-pixel mapping，class labels are allocated randomly to sub-pixels，and the location of each sub-pixel is optimized to obtain the final result6.The sub-pixel mapping methods based on the perimeter minimization，pixel swapping，and particle swarm optimization all belong to the initialization-then-optimization sub-pixel mapping7，8.However，this type usually has a long processing time due to the high computational complexity of optimization.On the other hand，the soft-then-hard sub-pixel mapping type has simple processing，which makes this type be more popular than the other sub-pixel mapping type9.The soft-then-hard type involves two steps，the sub-pixel sharpening and the class allocation.Fine fractional images with the land-cover class proportions corresponding to the sub-pixels are first obtained by sub-pixel sharpening using the methods based on super-resolution reconstruction，backpropagation neural network，spatial attraction model，indicator-cokriging，and Hopfield neural networks10-13.Then，the class labels are allocated to all sub-pixels by class allocation according to the proportions.The class allocation methods include the simulated annealing，the linear optimization，the highest fraction value，and units of classes14，15.Most soft-then-hard sub-pixel mapping methods are based on the spatial dependence assumption，namely，the closer the spatial distance is，the more likely the sub-pixels belong to the same land-cover class16-19.According to the spatial dependence assumption，Mertens et al.obtained the pixel-scale spatial information by using the spatial attraction model between sub-pixels and pixels，and the mapping result was derived according to this spatial information20.In order to obtain more precise scale spatial information，Ling et al.proposed a spatial attraction model between sub-pixels to obtain sub-pixel scale spatial information，improving the mapping accuracy of land-cover classes21.Chen et al.used the image segmentation algorithm to segment the abundance images，and then calculated the irregular scale areas from the segmentation images to obtain a kind of object-scale spatial information22.Further，Wang et al.used the random walk algorithm to consider the spatial dependence among and within the irregular scale areas at the same time，so as to obtain the better mapping results23.However，the current sub-pixel mapping methods usually use the dependence between sub-pixels in the specified rectangular local window to obtain the spectral information，as shown in Fig.1（a），and the number of spectral bands used is also little.But the distribution area of land-cover class is irregular in the actual environment，as shown in Fig.1（b），and the spectral information in each band is also different.Therefore，the spectral information in the current sub-pixel mapping methods is usually not accurate enough，affecting the final mapping result.To solve this issue，sub-pixel mapping based on spectral information of irregular scale areas（SIISA）for hyperspectral images is proposed in this paper.The contributions of this work are as follows：（1）Through establishing the normalized model，the proposed SIISA considers the spectral information of irregular scale areas and utilizes the spectral information of all bands，improving the accuracy of mapping results.（2）The proposed SIISA combines the spectral information of irregular scale areas with the spatial information of irregular scale areas generated from our previous work23 to obtain more accurate spatial-spectral information.The spatial-spectral information is closer to the real distribution of land-cover classes，which improves the performance of sub-pixel mapping.（3）The superiority of SIISA over the existing sub-pixel mapping methods is demonstrated by testing three remote sensing images.Fig.1Spatial information in（a）the rectangular local window and（b）the irregular scale areas图1空间信息（a）矩形局部窗口和（b）不规则尺度区域539红 外 与 毫 米 波 学 报42 卷This paper is organized as follows.Section I introduces the proposed method in detail.Section II shows the experimental analysis.Section III gives the conclusions.1 Method The overall process of SIISA is shown in Fig.2.The coarse original hyperspectral image is first upsampled by bicubic interpolation according to factor scale.The upsampled image is then unmixed and segmented to obtain the abundance image of each class and the segmentation image，respectively.Next，the random walker algorithm is used to calculate the proportional values of sub-pixels belonging to irregular scale areas from the fusion result of the abundance image of each class and segmentation image to obtain the spatial information of irregular scale areas.In addition，the normalized model is constructed to calculate the segmentation image to yield the spectral information of irregular scale areas for all bands.Finally，according to fusion results of spatial information and spectral information，class labels are assigned to sub-pixels to obtain the mapping result by class allocation method.Suppose the coarse original hyperspectral image is Y.The upsampled image Y is obtained by bicubic interpolation.The abundance image from the upsampled image Y is Hm（m=1，2，.，M，Mis the number of land-cover classes）with the proportional value Hm(pa)of sub-pixel pa（a=1，2，.，NS2，N is the number of mixed pixels，NS2 is the number of sub-pixels）belonging to the mthland-cover class.At the same time，the segmentation result from the upsampled image Y is Y with the irregular scale areas Oi（i=1，2，.，I，I is the number of irregular scale areas）by a segmentation scale parameter V，where Oi contains Ki sub-pixels.We integrate the abundance image of each class with the principal component of segmentation images to obtain the proportional values of sub-pixels in irregular scale areas.Therefore，the proportion value Um(Oi)of the irregular scale areas Oi belonging to the mthland-cover class is obtained by averaging these proportion values Hm(pa)of sub-pixels pa in this area，as shown in Eq.（1）.Um(Oi)=i=1KiHm()paKi.（1）Next，we will introduce in detail the two modules included in the proposed SIISA method，namely the spatial information module and the spectral information module.1.1Spatial information moduleIn spatial information module，we calculate the proportion value Um(Oi)of the irregular scale areas to obtain the spatial information Espa of irregular scale areas by using the random walk algorithm23，as shown in Eq.（2）.Espa=min m=1M(1-)Ewithinm(Um)+Eamongm()Um,（2）where Um=Um(O1)，Um(O2)，.，Um(Oi)is the column vector and is the empirical weight parameter，which is set to 0.5 here.Ewithinm(U)represents the internal spatial information of each irregular scale area，and Eamongm(U)represents the spatial information between adjacent irregular scale areas.They can be calculated by Eqs.（3）and（4），respectively.Ewithinm(Um)=n=1,n mMUTnnUn+(Um-1)Tm(Um-1),（3）Eamongm(Um)=UTmLUm,（4）where n is a diagonal matrix，where the value on the diagonal is the proportional value of each irregular scale area belonging to the nthland-cover class，and the value on the diagonal in m is the proportional value of each irregular scale area belonging to the mthland-cover class.The representation of 1 is a vector whose elements are 1.L is a Laplace matrix which represents the difference between adjacent areas，as shown in Eq.（5）.L=-zjq if j=q-zjq if j and q are adjacent areas 0 otherwise ,（5）where zjq=exp(-(yj-yq)2)is the spectral value difference between the jth irregular scale area Oj and the qth irregular scale area Oq.1.2Spectral information moduleIn spectral information module，the spectral information Espe of all bands in the irregular scale areas is obFig.2The flowchart of SIISA图2SIISA流程图5404 期 WANG Peng et al：Sub-pixel mapping based on spectral information of irregular scale areas for hyperspectral imagestained by using the previously obtained segmentation image Y.The segmentation image contains I irregular scale regions Oi，and each Oi includes Ki sub-pixels.Assuming that the spectrum of sub-pixels in each irregular scale area follows an approximate normal distribution24，a normalized model is constructed to calculate the spectral information of all bands in irregular scale areas as：Espe=min i=1I1KiBk=1Kij=1B()xk,j-xi,ji.j2,（6）where B is the number of spectral bands，and-xi，j and i.j is the average value and standard deviation of the spectral reflectance of the irregular scale area Oi in the band j.They are obtained by calculating the spectral reflectance of all sub-pixels in this irregular scale area.xk，j represents the spectral reflectance of the kth sub-pixel in the jth band in the irregular scale area Oi.The spatial information Espa and spectral information Espe are then integrated through the weight parameter to obtain the irregular scale spatial-spectral information E，as shown in Eq.（7）.E=min Espe+(1-)Espa.（7）Finally，the class allocation based on particle swarm optimization25 is used to optimize the objective function E to obtain the final mapping result.First，land-cover class labels are randomly assigned to all sub-pixels.Then，the labels of these sub-pixels are updated iteratively until the objective function reaches the minimum value.In each iteration，when the label assigned to a sub-pixel is converted to other labels，if the objective function value decreases，the conversion will be accepted，and if it increases，the conversion will be rejected.It is stipulated that when the converted sub-pixel is less than 0.1%of the total number，the class allocation terminates，obtaining the final mapping result.2 Experiment 2.1Experimental datasetThree datasets are tested to evaluate the performance of the proposed SIISA.According to the general experimental process of sub-pixel mapping，the original fine hyperspectral image is downsampled by an S S mean filter to obtain the simulated coarse image as input26，27.Due to its good robustness，the spectral unmixing method based on support vector machine is used to obtain the abundance images from the simulated coarse image10.A reference image is yielded by classifying the fine hyperspectral image.The weight parameter is selected as 0.6，0.6 and 0.5 for experiments 1，2 and 3，respectively.The segmentation scale parameter V is set to 10，10 and 5 for the three datasets，respectively.In the experiment 1，the performance of the proposed method is tested in the dataset from the multispec