Mohammad Reza Mobasheri

Unsupervised Feature Selection Using Geometrical Measures in Prototype Space for Hyperspectral Imagery

Feature/band selection is a common technique to overcome the “curse of dimensionality” posed by the high dimensionality of hyperspectral imagery. When the image is characterized by unknown phenomena, an unsupervised approach can be utilized to select the most distinctive and informative bands. The efficiency of an unsupervised feature selection (FS) depends on the criteria to be optimized and the space (e.g., feature space, pixel space, spectral space, etc.) in which the data are represented. Moreover, the determination of the initial feature and the determination of the optimal feature size (the optimal number of distinct bands to be selected) are other challenges faced in unsupervised approaches. In this paper, we propose two unsupervised FS methods by representing bands in the prototype space (PS). The first method proposes a way for selecting the initial feature based on the orthogonal distance from the PS diagonal and determines the optimal feature size by employing the HySime algorithm in the PS. The second method uses two criteria defined by the tangent of the angles between the band vectors in the PS in order to select the initial feature and to describe the band correlations. Meanwhile, the determination of the optimal feature size is embedded in this method. The experimental results on real and synthetic data sets show that our methods are more reliable and can yield a better result in terms of class separability and Friedman test than other widely used techniques.

Endmember Extraction Using a Combination of Orthogonal Projection and Genetic Algorithm

Common endmember extraction algorithms presume that the number of materials present is either known or may be predetermined by using spectral databases or other approaches. In this letter, we propose a new method called genetic orthogonal projection (GOP) for endmember extraction in imaging spectrometry. GOP is based on a fully unsupervised approach and uses convex geometric characteristics as well as a genetic algorithm. We compare GOP with existing endmember extraction algorithms and demonstrate that GOP partially outperforms them, without the need of a priori information.

A parametric method for estimation of leaf area index using landsat ETM+ data

Leaf Area Index (LAI) is a key variable for monitoring biophysical and biochemical characteristic of vegetation. So far, various remote-sensing methods are proposed to assess this index; each has its own advantages and limitations. In this study, the Scatterplot of NIR and Red bands (SNIR-R) of ETM+ images was used for LAI estimation. For this, nine different parameters consisting of five distances and four angles were extracted from SNIR-R. All possible combinations of these nine parameters were taken into account and as a result, 511 different regression equations were developed for estimation of LAI. The best regression equation (5P-LAI3) was made of two angles and three distances had the highest correlation coefficient (R) of 0.94 and root mean square (RMSE) of 0.75. On another approach, the triangle of scattered data in the SNIR-R was divided into three separate regions based on PVI (Perpendicular Vegetation Index) values. Three different regression equations were fitted to each region. Use of this Triangle Segmentation Model (TSM) improved the results slightly; that is, comparing with the results of general model 5P-LAI3, RMSE reduced to 0.66 and R increased to 0.96. The data collected throughout BigFoot project was used in this study. Comparing with other models in which BigFoot data were used, it was concluded that despite the simplicity of 5P-LAI3 model, it has an acceptable accuracy and TSM showed the highest accuracy, after all.

An Investigation Into Machine Learning Regression Techniques for the Leaf Rust Disease Detection Using Hyperspectral Measurement

The complex impacts of disease stages and disease symptoms on spectral characteristics of the plants lead to limitation in disease severity detection using the spectral vegetation indices (SVIs). Although machine learning techniques have been utilized for vegetation parameters estimation and disease detection, the effects of disease symptoms on their performances have been less considered. Hence, this paper investigated on 1) using partial least square regression (PLSR), v support vector regression (v-SVR), and Gaussian process regression (GPR) methods for wheat leaf rust disease detection, 2) evaluating the impact of training sample size on the results, 3) the influence of disease symptoms effects on the predictions performances of the above-mentioned methods, and 4) comparisons between the performances of SVIs and machine learning techniques. In this study, the spectra of the infected and non infected leaves in different disease symptoms were measured using a non imaging spectroradiometer in the electromagnetic region of 350 to 2500 nm. In order to produce a ground truth dataset, we employed photos of a digital camera to compute the disease severity and disease symptoms fractions. Then, different sample sizes of collected datasets were utilized to train each method. PLSR showed coefficient of determination (R2) values of 0.98 (root mean square error (RMSE) = 0.6) and 0.92 (RMSE = 0.11) at leaf and canopy, respectively. SVR showed R2 and RMSE close to PLSR at leaf (R2 = 0.98, RMSE = 0.05) and canopy (R2 = 0.95, RMSE = 0.12) scales. GPR showed R2 values of 0.98 (RMSE = 0.03) and 0.97 (RMSE = 0.11) at leaf and canopy scale, respectively. Moreover, GPR represents better performances than others using small training sample size. The results represent that the machine learning techniques in contrast to SVIs are not sensitive to different disease symptoms and their results are reliable.

Soil moisture content assessment based on Landsat 8 red, near-infrared, and thermal channels

Abstract. Soil moisture content (SMC) plays an important role in different environmental. In this study, four different soil moisture indices, namely, SOMID, SOMID-FS, SOMID-FT, and CSOMID-FT, were introduced. In this work, the following parameters were used to estimate SMC at a depth of 5 cm: (a) the distance of pixels from the origin in the scatter-plot of near-infrared (NIR) and red bands (SNIR-R), (b) the fraction of soil cover in each pixel, and (c) the land surface temperature. It was concluded that the CSOMID-FT was the most accurate index for estimation of SMC (RMSE=0.045, R=0.92). This index divides the SNIR-R into three separate regions based on the pixels’ normalized difference vegetation index (NDVI) values and assigns a specific regression equation to each region. The results showed that as the NDVI values increase, the accuracy of the proposed indices decreases. Furthermore, the SOMID-FT and CSOMID-FT were used to estimate SMC at five different depths of 5, 10, 20, 50, and 100 cm. It was concluded that the satellite-estimated SMC was highly correlated with the field-measured data at 5-cm soil depth.

Leaf Equivalent Water Thickness assessment using reflectance at optimum wavelengths

Leaf water content is an important parameter in environmental monitoring. The present study investigated the relation between leaf Equivalent Water Thickness (EWT) as a parameter to estimate the leaf water content and the reflectance in 400-2,500 nm spectral range. The data used were the well-known Leaf Optical Properties Experiment 93 (LOPEX93) field collected data. Four hundred leaf samples were used, 320 of which for modelling and the remaining 80 for testing the model. Four different approaches were investigated in this study: 1) linear regression between reflectance in individual wavelength and EWT; 2) the difference of reflectance in two wavelengths and EWT; 3) ratio of reflectance in two wavelengths and EWT; and finally 4) the normalized difference of reflectance in two different wavelengths and EWT. The results showed that the band combinations such as ratio and normalized difference had higher regressions with leaf water content. In addition, the findings of this study showed that some parts of the near infrared (NIR) and short wave infrared (SWIR) of the spectrum provided higher accuracies in EWT assessment, and correlations of more than 90% were achieved. Finally, this investigation showed that a wide range of wavelengths could be used for EWT assessment task. Despite the general belief in using water absorption bands for leaf water content assessment, this study shows that water absorption bands are not necessarily productive as other wavelengths have the potential to generate better results.

Windowed Fourier Transform for Noise Reduction of SAR Interferograms

In this letter, since these methods are able to process signals locally, two spatial frequency analyses including windowed Fourier transform and wavelet transform are used to reduce synthetic aperture radar interferometric phase noise.

Clustering Multispectral Images Using Spatial–Spectral Information

Clustering is an important topic in image analysis and has many applications. Owing to the limitations of the feature space in multispectral images and spectral overlap of the clusters, it is required to use some additional information such as the spatial context in image clustering. To increase the accuracy of image clustering, a new Hierarchical Iterative Clustering Algorithm using Spatial and Spectral information (HICLASS) is introduced. This algorithm separates pixels into uncertain and certain categories based on decision distances in the feature space. The algorithm labels the certain pixels using the k-means clustering, and the uncertain ones with the help of information in both spatial and spectral domains of the image. The proposed algorithm is tested using simulated and real data. The benchmark results indicate better performance of HICLASS when compared with the k-means, local embeddings, and some proximity-based algorithms. The overall accuracy of the k-means has increased between 12.5% and 20.4% for different data. The HICLASS method increases the accuracy and generates more homogeneous regions, which are required for object-based applications.

Classification by diagnosing all absorption features (CDAF) for the most abundant minerals in airborne hyperspectral images

Imaging through hyperspectral technology is a powerful tool that can be used to spectrally identify and spatially map materials based on their specific absorption characteristics in electromagnetic spectrum. A robust method called Tetracorder has shown its effectiveness at material identification and mapping, using a set of algorithms within an expert system decision-making framework. In this study, using some stages of Tetracorder, a technique called classification by diagnosing all absorption features (CDAF) is introduced. This technique enables one to assign a class to the most abundant mineral in each pixel with high accuracy. The technique is based on the derivation of information from reflectance spectra of the image. This can be done through extraction of spectral absorption features of any minerals from their respected laboratory-measured reflectance spectra, and comparing it with those extracted from the pixels in the image. The CDAF technique has been executed on the AVIRIS image where the results show an overall accuracy of better than 96%.

Exploring the use of satellite images in the estimation of potential malaria outbreak regions

Malaria outbreaks affect nearly 40 per cent of the earths population, most of whom live in tropical and subtropical zones. Malaria is an infectious disease that is transferred by the female mosquito of the species Anopheles. The life cycle of the malaria parasite develops in the anopheline and in the human body. These parasites require suitable environmental conditions in order to complete their development cycles within the mosquito. The relevant parameters are temperature, humidity, vegetation and water. A temperature range of 25–35°C and relative humidity range of 50–80 per cent is suitable for developing malaria outbreaks. As the fly-range of the mosquito is limited to 2–4 km, and since water pools are necessary for breeding, the vector abundance is significantly higher around water bodies. Vegetation cover also has an indirect role on malaria vector abundance. To seek to locate the regions with high potential for malaria outbreaks, we have constructed an experimental map of above-mentioned parameters via remote sensing images. A 7ETM+ image of Landsat platform is used in this study and maps of parameters such as land surface temperature, air temperature, air humidity, water pools and vegetated area were produced. However, a weighted combination of these layers showed some poor agreement with the distribution of positive malaria cases collected in the health centres in the region. The methodology first developed in this study is fast and accurate enough to be relied on for forecasting purposes, and could eventually lead—after further research and proper correlations—to improving the targeting of mitigation and relief operations by local health and related organizations.