Massimo Zanetti

Sequential Spectral Change Vector Analysis for Iteratively Discovering and Detecting Multiple Changes in Hyperspectral Images

This paper presents an effective semiautomatic method for discovering and detecting multiple changes (i.e., different kinds of changes) in multitemporal hyperspectral (HS) images. Differently from the state-of-the-art techniques, the proposed method is designed to be sensitive to the small spectral variations that can be identified in HS images but usually are not detectable in multispectral images. The method is based on the proposed sequential spectral change vector analysis, which exploits an iterative hierarchical scheme that at each iteration discovers and identifies a subset of changes. The approach is interactive and semiautomatic and allows one to study in detail the structure of changes hidden in the variations of the spectral signatures according to a top-down procedure. A novel 2-D adaptive spectral change vector representation (ASCVR) is proposed to visualize the changes. At each level this representation is optimized by an automatic definition of a reference vector that emphasizes the discrimination of changes. Finally, an interactive manual change identification is applied for extracting changes in the ASCVR domain. The proposed approach has been tested on three hyperspectral data sets, including both simulated and real multitemporal images showing multiple-change detection problems. Experimental results confirmed the effectiveness of the proposed method.

Rayleigh-Rice Mixture Parameter Estimation via EM Algorithm for Change Detection in Multispectral Images

The problem of estimating the parameters of a Rayleigh-Rice mixture density is often encountered in image analysis (e.g., remote sensing and medical image processing). In this paper, we address this general problem in the framework of change detection (CD) in multitemporal and multispectral images. One widely used approach to CD in multispectral images is based on the change vector analysis. Here, the distribution of the magnitude of the difference image can be theoretically modeled by a Rayleigh-Rice mixture density. However, given the complexity of this model, in applications, a Gaussian-mixture approximation is often considered, which may affect the CD results. In this paper, we present a novel technique for parameter estimation of the Rayleigh-Rice density that is based on a specific definition of the expectation-maximization algorithm. The proposed technique, which is characterized by good theoretical properties, iteratively updates the parameters and does not depend on specific optimization routines. Several numerical experiments on synthetic data demonstrate the effectiveness of the method, which is general and can be applied to any image processing problem involving the Rayleigh-Rice mixture density. In the CD context, the Rayleigh-Rice model (which is theoretically derived) outperforms other empirical models. Experiments on real multitemporal and multispectral remote sensing images confirm the validity of the model by returning significantly higher CD accuracies than those obtained by using the state-of-the-art approaches.

A generalized statistical model for binary change detection in multispectral images

Recently, a thresholding method based on the Rayleigh-Rice mixture has been proposed for solving binary change detection problems in multispectral image pairs. However, when images acquired by the last generation of multispectral scanners having high radiometric resolution are considered, the distribution fitting is still not satisfactory and computed thresholds remain quite distant from the optimal values. The main reason for this seems to be that in all previous approaches the unchange class is modeled as a single class. Instead, both practice and recent studies showed that this is not the case for new generation data. In this work, we propose a generalized statistical model for the difference image that allows the unchange class to be complex. The resulting model has more degrees of freedom, therefore it better fits real distributions and returns almost optimal thresholds for binary decision also with high radiometric resolution images.

Serial and parallel approaches for image segmentation by numerical minimization of a second-order functional

Abstract Because of its attractive features, second order segmentation has shown to be a promising tool in remote sensing. A known drawback about its implementation is computational complexity, above all for large set of data. Recently in Zanetti et al. [1], an efficient version of the block-coordinate descent algorithm (BCDA) has been proposed for the minimization of a second order elliptic approximation of the Blake–Zissermann functional. Although the parallelization of linear algebra operations is expected to increase the performance of BCDA when addressing the segmentation of large-size gridded data (e.g., full-scene images or Digital Surface Models (DSMs)), numerical evidence shows that this is not sufficient to get significant reduction of computational time. Therefore a novel approach is proposed which exploits a decomposition technique of the image domain into tiles. The solution can be computed by applying BCDA on each tile in parallel way and combining the partial results corresponding to the different blocks of variables through a proper interconnection rule. We prove that this parallel method (OPARBCDA) generates a sequence of iterates which converges to a critical point of the functional on the level set devised by the starting point. Furthermore, we show that the parallel method can be efficiently implemented even in a commodity multicore CPU. Numerical results are provided to evaluate the efficiency of the parallel scheme on large images in terms of computational cost and its effectiveness with respect to the behavior on the tile junctions.

A tiling procedure for second-order variational segmentation of large size remote sensing images

Typical tiling approaches to segmentation of large images perform separated runs of a specific segmentation algorithm on tiles and then merge the results. However, specific post-processing is often required to remove possible artifacts on tiles junctions. In this paper, we aim at showing that a simple tiling strategy with partially overlapping tiles can be applied to a 2-nd order variational segmentation method based on the minimization of the Blake-Zisserman functional, in such a way that tile boundaries are coherent without any need of specific post-processing. Moreover, the energy minimization is performed on each tile with Dirichlet initial boundary conditions; thus, tiles are independent and the whole procedure is parallelizable with independent tiles.

Piecewise Linear Approximation of Vector-Valued Images and Curves via Second-Order Variational Model

Variational models are known to work well for addressing image restoration/regularization problems. However, most of the methods proposed in the literature are defined for scalar inputs and are used on multiband images (such as RGB or multispectral imagery) by the composition of a simple band-wise processing. This involves suboptimal results and may introduce artifacts. Only in a few cases, variational models are extended to the case of vector-valued inputs. However, the known implementations are restricted to the first-order models, while the second-order models are never considered. Thus, typical problems of the first-order models, such as the staircasing effect cannot be overtaken. This paper considers a second-order functional model to function approximation with free discontinuities given by Blake–Zisserman (BZ) and proposes an efficient minimization algorithm in the case of vector-valued inputs. In the BZ model, the Hessian of the solution is penalized outside a set of finite length, therefore the solution is forced to be piecewise linear. Moreover, the model allows the formation of free discontinuities and free gradient discontinuities. The proposed algorithm is applied to difficult color image restoration/regularization problems and to piecewise linear approximation of curves in space.

Analysis of multitemporal Sentinel-2 images in the framework of the ESA Scientific Exploitation of Operational Missions

This paper focuses on the scientific preliminary results of the project “S2-4Sci Land and Water - Multitemporal Analysis” funded by the European Space Agency (ESA) in the framework of the Scientific Exploitation of Operational Missions (SEOM). The aim of the project is the development of advanced multitemporal methods tailored on the specific properties of S2 images. The Sentinel-2 (S2) constellation has a huge potential for multitemporal analysis, due to the increased geometrical resolution, the novel spectral capabilities, a swath width of 290Km and the short revisit time. Three main applications and methodological areas are investigated: i) land-cover maps updating, ii) land-cover change detection, and iii) time series analysis. The proposed approaches are briefly described and preliminary results obtained on S2 images are discussed.

A parallel approach for image segmentation by numerical minimization of a second-order functional

Because of its attractive features, image segmentation has shown to be a promising tool in remote sensing. A known drawback about its implementation is computational complexity. Recently in [1] an effcient numerical method has been proposed for the minimization of a second-order variational approximation of the Blake-Zissermann functional. The method is an especially tailored version of the block-coordinate descent algorithm (BCDA). In order to enable the segmentation of large-size gridded data, such as Digital Surface Models, we combine a domain decomposition technique with BCDA and a parallel interconnection rule among blocks of variables. We aim to show that a simple tiling strategy enables us to treat large images even in a commodity multicore CPU, with no need of specific post-processing on tiles junctions. From the point of view of the performance, little computational effort is required to separate data in subdomains and the running time is mainly spent in concurrently solving the independent subpr...

Edge-crease detection and surface reconstruction from point clouds using a second-order variational model

The automatic detection of geometric features, such as edges and creases, from objects represented by 3D point clouds (e.g., LiDAR measurements, Tomographic SAR) is a very important issue in different application domains including urban monitoring and building reconstruction. A limitation of many methods in the literature is that they rely on rasterization or interpolation of the original grid, with consequent potential loss of detail. Recently, a second-order variational model for edge and crease detection and surface regularization has been presented in literature and succesfully applied to DSMs. In this paper we address the generalization of this model to unstructured grids. The model is based on the Blake-Zisserman energy and allows to obtain a regularization of the original data (noise reduction) which does not affect crucial regions containing jumps and creases. Specifically, we focus on the detection of these features by means of two auxiliary functions that are computable by solving specific differential equations. Results obtained on LiDAR data by solving the equations via Finite Element Method are presented.