图像滤波算子在土壤盐渍化信息提取中的应用

王筝; 张飞; 张贤龙; 王一山

doi:doi:10.3788/LOP57.042801

激光与光电子学进展, 2020, 57 (4): 042801, 网络出版: 2020-02-20

图像滤波算子在土壤盐渍化信息提取中的应用下载： 954次

Application of Image Filtering Operator in Extraction of Soil Salinization Information

王筝 ^1,2张飞 ^1,2,3,*张贤龙 ^1,2王一山 ^1,2

作者单位

¹ 新疆大学资源与环境科学学院智慧城市与环境建模自治区普通高校重点实验室, 新疆乌鲁木齐 830046

² 新疆大学绿洲生态教育部重点实验室, 新疆乌鲁木齐 830046

³ 中亚地理信息开发利用国家测绘地理信息局工程技术研究中心, 新疆乌鲁木齐 830046

图 & 表

图 1. 研究区示意图。(a)艾比湖湿地保护区及采样点分布图;(b)艾比湖湿地保护区内盐渍化土壤;(c)艾比湖保护区内水塘表面盐结晶;(d)艾比湖保护区内植被

Fig. 1. Schematic of research area. (a) Ebinur Lake Wetland Reserve and distribution of sampling points; (b) salinized soil in Ebinur Lake Wetland Reserve; (c) salt crystals on surface of ponds in Ebinur Lake Wetland Reserve; (d) vegetation in Ebinur Lake Wetland Reserve

下载图片查看原文

图 2. 采样点盐渍化程度专题图

Fig. 2. Matic map of salinization degree of sampling points

下载图片查看原文

图 3. 支持向量机分类过程示意图

Fig. 3. Schematic of classification process of SVMs

下载图片查看原文

图 4. 经不同滤波方法处理后的影像。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

Fig. 4. Images processed by different filtering methods. (a) Raw remote-sensing image; (b) Laplacian filtering; (c) high-pass filtering; (d) low-pass filtering; (e) Gaussian high-pass filtering; (f) Gaussian low-pass filtering; (g) median filtering; (h) directional filtering

下载图片查看原文

图 5. 不同波段的亮度值变化图

Fig. 5. Variation in brightness value at different bands

下载图片查看原文

图 6. 基于不同滤波方法处理的遥感影像的分类图。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

Fig. 6. Classification of remote-sensing images based on different filtering methods. (a) Raw remote-sensing image; (b) Laplacian filtering; (c) high-pass filtering; (d) low-pass filtering; (e) Gaussian high-pass filtering; (f) Gaussian low-pass filtering; (g) median filtering; (h) directional filtering

下载图片查看原文

表 1土壤盐渍化程度分级

Table1. Classification of degree of soil salinization

Degree ofsoil salinization	Soil saltcontent /(g·kg^-1)	Numberof samples	Growth condition
Non-saline soil	<1	10	Healthy growth of vegetation
Mildly saline soil	1--6	15	Plant coverage is approximately 15% to 30%,and salt-sensitive vegetation may be affected
Moderately saline soil	6--10	6	Plant coverage isapproximately 10% to 15%,and salt-tolerant crops are less affected
Severely saline soil	10--20	6	Plant coverage is approximately 5% to 10%,and salt-tolerant crops and their yields are greatly affected
Saline soil	>20	1	There is only a small amount of salt-tolerantvegetation such as Haloxylon ammodendron

查看原文

表 2滤波矩阵函数

Table2. Filtering matrix functions

Filtering	Filtering matrix
Laplacian	$\begin{matrix} \begin{matrix} 0 & - 1 & 0 \\ - 1 & 4 & - 1 \\ 0 & - 1 & 0 \end{matrix} \end{matrix}$
High-pass	$\begin{matrix} \begin{matrix} - 1 & - 1 & - 1 \\ - 1 & 8 & - 1 \\ - 1 & - 1 & - 1 \end{matrix} \end{matrix}$
Low-pass	$\begin{matrix} \begin{matrix} f (i - 1, j + 1) & f (i, j + 1) & f (i + 1, j + 1) \\ f (i - 1, j) & f (i, j) & f (i + 1, j) \\ f (i - 1, j - 1) & f (i, j - 1) & f (i + 1, j - 1) \end{matrix} \end{matrix}$
Gaussianhigh-pass	$\begin{matrix} \begin{matrix} - 0.0007 & - 0.0256 & - 0.0007 \\ - 0.0256 & 0.1025 & - 0.0256 \\ - 0.0007 & - 0.0256 & - 0.0007 \end{matrix} \end{matrix}$
Gaussianlow-pass	$\begin{matrix} \begin{matrix} 0.0007 & 0.0256 & 0.0007 \\ 0.0256 & 0.8948 & 0.0256 \\ 0.0007 & 0.0256 & 0.0007 \end{matrix} \end{matrix}$
Median	$\begin{matrix} \begin{matrix} f (i - 1, j + 1) & f (i, j + 1) & f (i + 1, j + 1) \\ f (i - 1, j) & f (i, j) & f (i + 1, j) \\ f (i - 1, j - 1) & f (i, j - 1) & f (i + 1, j - 1) \end{matrix} \end{matrix}$
Directional	$\begin{matrix} \begin{matrix} 0 & - 1 & 1 \\ - 1 & 0 & - 1 \\ 1 & - 1 & 0 \end{matrix} \end{matrix}$

查看原文

表 3研究区遥感影像分类方案

Table3. Classification scheme of remote-sensing image of research area

Classificationtype	Soil saltcontent /(g·kg^-1)	Description
Water body	-	The color of the water on thefalse color image is blue or black,including rivers, ditches, lakes, etc.
Non-saline soil	<1	The soil has low salt content andlow image reflectance, includingrocks, wasteland, mountains, etc.
Mildly saline soil	1--6	The soil has less salt content, and thevegetation coverage is about 8%~15%The white patches in the middle of thevegetation and the bright spot area are small
Moderately saline soil	6--10	The soil has a general salt content,vegetation coverage is about 1% to 8%, andthere are fewer white patches on the image
Severely saline soil	10--20	The soil has a high salt contentand is a heavily salinizedarea in the Ebinur Lake region
Saline soil	>20	The soil has high salt content, high spectralreflectance, obvious salt crust on the surface,white plaque distribution on the image,and basically no vegetation growth

查看原文

表 4支持向量机的分类精度评价方法

Table4. Assessment methods of classification accuracy of SVM

Filtering	Overallaccuracy /%	Kappacoefficient /%
Raw remotesensing image	86.7285	82.21
High-pass	87.0441	82.65
Low-pass	89.6555	86.15
Laplacian	88.6544	84.80
Directional	88.8714	85.10
Gaussian high-pass	89.6950	86.20
Gaussian low-pass	89.6087	86.58
Median	89.6678	86.16

查看原文

王筝, 张飞, 张贤龙, 王一山. 图像滤波算子在土壤盐渍化信息提取中的应用[J]. 激光与光电子学进展, 2020, 57(4): 042801. Zheng Wang, Fei Zhang, Xianlong Zhang, Yishan Wang. Application of Image Filtering Operator in Extraction of Soil Salinization Information[J]. Laser & Optoelectronics Progress, 2020, 57(4): 042801.

图像滤波算子在土壤盐渍化信息提取中的应用下载： 954次

图 1. 研究区示意图。(a)艾比湖湿地保护区及采样点分布图;(b)艾比湖湿地保护区内盐渍化土壤;(c)艾比湖保护区内水塘表面盐结晶;(d)艾比湖保护区内植被

Fig. 1. Schematic of research area. (a) Ebinur Lake Wetland Reserve and distribution of sampling points; (b) salinized soil in Ebinur Lake Wetland Reserve; (c) salt crystals on surface of ponds in Ebinur Lake Wetland Reserve; (d) vegetation in Ebinur Lake Wetland Reserve

图 2. 采样点盐渍化程度专题图

Fig. 2. Matic map of salinization degree of sampling points

图 3. 支持向量机分类过程示意图

Fig. 3. Schematic of classification process of SVMs

图 4. 经不同滤波方法处理后的影像。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

Fig. 4. Images processed by different filtering methods. (a) Raw remote-sensing image; (b) Laplacian filtering; (c) high-pass filtering; (d) low-pass filtering; (e) Gaussian high-pass filtering; (f) Gaussian low-pass filtering; (g) median filtering; (h) directional filtering

图 5. 不同波段的亮度值变化图

Fig. 5. Variation in brightness value at different bands

图 6. 基于不同滤波方法处理的遥感影像的分类图。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

表 1土壤盐渍化程度分级

Table1. Classification of degree of soil salinization

表 2滤波矩阵函数

Table2. Filtering matrix functions

表 3研究区遥感影像分类方案

Table3. Classification scheme of remote-sensing image of research area

表 4支持向量机的分类精度评价方法

Table4. Assessment methods of classification accuracy of SVM

关于本站 Cookie 的使用提示

全站搜索

图像滤波算子在土壤盐渍化信息提取中的应用 下载： 954次

图 1. 研究区示意图。(a)艾比湖湿地保护区及采样点分布图;(b)艾比湖湿地保护区内盐渍化土壤;(c)艾比湖保护区内水塘表面盐结晶;(d)艾比湖保护区内植被

Fig. 1. Schematic of research area. (a) Ebinur Lake Wetland Reserve and distribution of sampling points; (b) salinized soil in Ebinur Lake Wetland Reserve; (c) salt crystals on surface of ponds in Ebinur Lake Wetland Reserve; (d) vegetation in Ebinur Lake Wetland Reserve

图 2. 采样点盐渍化程度专题图

Fig. 2. Matic map of salinization degree of sampling points

图 3. 支持向量机分类过程示意图

Fig. 3. Schematic of classification process of SVMs

图 4. 经不同滤波方法处理后的影像。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

Fig. 4. Images processed by different filtering methods. (a) Raw remote-sensing image; (b) Laplacian filtering; (c) high-pass filtering; (d) low-pass filtering; (e) Gaussian high-pass filtering; (f) Gaussian low-pass filtering; (g) median filtering; (h) directional filtering

图 5. 不同波段的亮度值变化图

Fig. 5. Variation in brightness value at different bands

图 6. 基于不同滤波方法处理的遥感影像的分类图。(a)原始遥感图像;(b)拉普拉斯滤波;(c)高通滤波;(d)低通滤波;(e)高斯高通滤波;(f)高斯低通滤波;(g)中值滤波;(h)方向滤波

表 1土壤盐渍化程度分级

Table1. Classification of degree of soil salinization

表 2滤波矩阵函数

Table2. Filtering matrix functions

表 3研究区遥感影像分类方案

Table3. Classification scheme of remote-sensing image of research area

表 4支持向量机的分类精度评价方法

Table4. Assessment methods of classification accuracy of SVM

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索

图像滤波算子在土壤盐渍化信息提取中的应用下载： 954次