Statistical Region Merging Algorithm Research Articles

Fast Line Segment Detection and Large Scene Airport Detection for PolSAR

In this paper, we propose a fast Line Segment Detection algorithm for Polarimetric synthetic aperture radar (PolSAR) data (PLSD). We introduce the Constant False Alarm Rate (CFAR) edge detector to obtain the gradient map of the PolSAR image, which tests the equality of the covariance matrix using the test statistic in the complex Wishart distribution. A new filter configuration is applied here to save time. Then, the Statistical Region Merging (SRM) framework is utilized for the generation of line-support regions. As one of our main contributions, we propose a new Statistical Region Merging algorithm based on gradient Strength and Direction (SRMSD). It determines the merging predicate with consideration of both gradient strength and gradient direction. For the merging order, we set it by bucket sort based on the gradient strength. Furthermore, the pixels are restricted to belong to a unique region, making the algorithm linear in time cost. Finally, based on Markov chains and a contrario approach, the false alarm control of line segments is implemented. Moreover, a large scene airport detection method is designed based on the proposed line segment detection algorithm and scattering characteristics. The effectiveness and applicability of the two methods are demonstrated with PolSAR data provided by UAVSAR.

A Region-Based Deep Reinforcement Learning Classification Method for GF-3 PolSAR Imagery Classification

The C-band GF-3 satellite is the first fully polarimetric synthetic aperture (PolSAR) designed by China, which has multi-polarization image modes and been used in various applications. Land cover classification is an effective approach in PolSAR image interpretation and further application. However, conventional classification methods are mostly pixel-based and are easily affected by inherent speckle noise. In addition, the feature selection of PolSAR image and the amount of training samples are also critical to the performance of classifiers. To solve these problems, in this paper, we propose a region-based PolSAR image classification method, which uses reinforcement learning method altogether with statistical region merging algorithm to improve the classification performance. The contributions of our method are mainly reflected in three aspects: First, the T3 matrix is considered as the only feature set in our method, including image segmentation and classification. Second, the region is produced via statistical region merging algorithm. Finally, a deep reinforcement learning model is used to obtain PolSAR image classification result. To evaluate the performance of the proposed method, two real GF-3 images are performed in the experiments, and the experimental results illustrate that the proposed method outperforms the conventional methods (support vector machine, random forest, and convolution neural network) in terms of accuracy and achieves the state-of-art results.

Building Change Detection Using a Shape Context Similarity Model for LiDAR Data

In this paper, a novel building change detection approach is proposed using statistical region merging (SRM) and a shape context similarity model for Light Detection and Ranging (LiDAR) data. First, digital surface models (DSMs) are generated from LiDAR acquired at two different epochs, and the difference data D-DSM is created by difference processing. Second, to reduce the noise and registration error of the pixel-based method, the SRM algorithm is applied to segment the D-DSM, and multi-scale segmentation results are obtained under different scale values. Then, the shape context similarity model is used to calculate the shape similarity between the segmented objects and the buildings. Finally, the refined building change map is produced by the k-means clustering method based on shape context similarity and area-to-length ratio. The experimental results indicated that the proposed method could effectively improve the accuracy of building change detection compared with some popular change detection methods.

Brain and pancreatic tumor segmentation using SRM and BPNN classification

As of late, to enhance the features of serviceability in medical clinic management, medical image processing plays progressive development in conditions of modus operandi and applications. Various techniques are used to diagnosis tumor parts in modern medical image processing with the rising demand in the respective field. In this paper, the detection of the brain tumor and pancreatic tumor using DBCWMF (Decision Based Couple Window Median Filter)algorithm, Statistical region merging (SRM), Cat Swarm Optimization and Scale-invariant feature transform (CSO-SIFT) extraction and classification through Back Propagation Neural Network (BPNN) is presented. DBCWMF works effectively in the preprocessing compared to Median and PGPD filter, segmentation done with SRM algorithm. After that, the feature selection techniques CSO and SIFT are used for detecting the part in tumor images which is affected and final classification through BPNN classification works effectively compared to ANN and AdaBoost classifier. The experimental tested on images from Medical Harvard School database and The Cancer Imaging Archive (TCIA) repository’s database.

Automatic Pulmonary Nodule Growth Measurement through CT Image Analysis based on Morphology Filtering and Statistical Region Merging

This paper proposes an innovative method for automatic detection of pulmonary nodules in Computed Tomography (CT) data and measurement of changes in the number and sizes of the detected nodules during the treatment session. In the presented method, two multi-slice CT images are first taken from the patient’s lung, each captured by a similar capturing device but at two different dates. The CT images are then analyzed and their pulmonary nodules are extracted using a novel framework based on Mathematical Morphology Filtering (MMF), Statistical Region Merging (SRM), and Support Vector Machines (SVM). The MMF step smoothes the image in order to increase its homogeneity as well as removing the noises and artifacts. The SRM algorithm segments each slice of the CT image. After connecting the boundaries of the segments in adjacent slices, three-dimensional objects are produced which are considered as nodule-candidates. These candidates are classified into nodules and non-nodules using a two-class SVM classifier. The extracted nodules in each image are then labeled and their characteristics (i.e. labels, locations, and sizes) are stored. Finally, after registering the image pair using an affine algorithm, the growth rates of the lung nodules are measured.

Towards semantic segmentation of orthophoto images using graph-based community identification

We present an unsupervised framework that automatically detects objects of interest in images by formulating the general problem of semantic segmentation as community detection problem in graphs. The framework broadly follows a four-step procedure. First, we perform an over-segmentation of the original image using the well-known statistical region merging algorithm which presents the advantage of not requiring any quantization or colour space transformations. Second, we compute the feature descriptors of the resulting segmented regions. For encoding colour and other textural information, each region is described by an hybrid descriptor based on colour histograms and covariance matrix descriptor. Third, from the set of descriptors we construct different weighted graphs using various graph construction algorithms. Finally, the resulting graphs are then divided into groups or communities using a community detection algorithm based on spectral modularity maximization. This algorithm makes use of the eigenspectrum of matrices such as the graph Laplacian matrix and the modularity matrix which are more likely to reveal the community structure of the graph. Experiments conducted on large orthophotos depicting several zones in the region of Belfort city situated on the north-eastern of France provide promising results. The proposed framework can be used by semi-automatic approaches to handle the challenging problems of scene parsing.

Segmentation of Retinal Nerve Fiber Layer in Optical Coherence Tomography (OCT) Images using Statistical Region Merging Technique for Glaucoma Screening

clinical parameter used to diagnose glaucoma eye disease. The thickness of the RNFL decreases as the intraocular pressure inside the eye increases. Decrease in RNFL thickness or damages to RNFL due to high intraocular pressure leads to Glaucoma. The present work provides a technique for the segmentation of Retinal Nerve Fiber Layer (RNFL) from the Optical Coherence Tomography (OCT) images. First, OCT image is pre-processed then RNFL in OCT is segmented by statistical region merging algorithm. The segmented RNFL is refined using morphological operations such as dilation and erosion. At the end, segmented RNFL is extracted by developed algorithm. The developed technique is tested with 50 RNFL images. The result shows the exact boundaries of RNFL and segmented portion of the RNFL. Segmented RNFL can be used to find the thickness of RNFL for detection of glaucoma.

Polarimetric SAR Image Segmentation Using Statistical Region Merging

The statistical region merging (SRM) algorithm exhibits efficient performance in solving significant noise corruption and does not depend on the data distribution. These advantages make SRM suitable for the segmentation of synthetic aperture radar (SAR) images, which are characterized by speckle noise and different distributions of various data types and spatial resolutions. However, the original SRM algorithm is designed for RGB and gray images characterized by additive noise and having a range of [0, 255]. In this letter, the SRM algorithm is generalized so that it can be applied to images with larger range and multiplicative noise. The original 4-neighborhood models are also generalized into 8-neighborhood models. The effectiveness of the generalized SRM (GSRM) algorithm is demonstrated by AirSAR and ESAR L-band Polarimetric SAR (PolSAR) data. Given that the input data of the GSRM algorithm can be single- or multi-dimensional, the proposed GSRM algorithm can be used for single- and multi-polarized as well as for fully polarimetric SAR data.

Segmentation of noisy colour images using Cauchy distribution in the complex wavelet domain

This study proposes a novel image segmentation technique for noisy colour images, in which the heavy-tailed characteristics of the image are modelled by Cauchy distributions. First, the RGB colour bands of the noisy image are decomposed into multiresolution representations using the dual-tree complex wavelet transform. For each wavelet subband, a model is built assuming that the input coefficients are contaminated with signal-independent additive white Gaussian noise. Hence, the authors derive an estimation rule in the wavelet domain to obtain the noise-free coefficients based on the bivariate Cauchy distribution. The bivariate model makes it possible to exploit the inter-scale dependencies of wavelet coefficients. Subsequently, the image is roughly segmented into textured and non-textured regions using the bivariate model parameters corresponding to the denoised coefficients. A multiscale segmentation is then applied to the resulting regions. Finally, a novel statistical region merging algorithm is introduced by measuring the Kullback-Leibler distance between the estimated Cauchy models for the neighbouring segments. The experiments demonstrate that the authors algorithm yields robust segmentation results for noisy images containing artificial or natural noise.

Segmentation-Driven Image Fusion Based on Alpha-Stable Modeling of Wavelet Coefficients

A novel region-based image fusion framework based on multiscale image segmentation and statistical feature extraction is proposed. A dual-tree complex wavelet transform (DT-CWT) and a statistical region merging algorithm are used to produce a region map of the source images. The input images are partitioned into meaningful regions containing salient information via symmetric alpha-stable (S alphaS) distributions. The region features are then modeled using bivariate alpha-stable (B alphaS) distributions, and the statistical measure of similarity between corresponding regions of the source images is calculated as the Kullback-Leibler distance (KLD) between the estimated B alphaS models. Finally, a segmentation-driven approach is used to fuse the images, region by region, in the complex wavelet domain. A novel decision method is introduced by considering the local statistical properties within the regions, which significantly improves the reliability of the feature selection and fusion processes. Simulation results demonstrate that the bivariate alpha-stable model outperforms the univariate alpha-stable and generalized Gaussian densities by not only capturing the heavy-tailed behavior of the subband marginal distribution, but also the strong statistical dependencies between wavelet coefficients at different scales. The experiments show that our algorithm achieves better performance in comparison with previously proposed pixel and region-level fusion approaches in both subjective and objective evaluation tests.