Ali Pour Yazdanpanah

Sky Segmentation by Fusing Clustering with Neural Networks

Sky segmentation is an important task for many applications related to obstacle detection and path planning for autonomous air and ground vehicles. In this paper, we present a method for the automated sky segmentation by fusing K-means clustering and Neural Network (NN) classifications. The performance of the method has been tested on images taken by two Hazcams (ie., Hazard Avoidance Cameras) on NASA’s Mars rover. Our experimental results show high accuracy in determining the sky area. The effect of various parameters is demonstrated using Receiver Operating Characteristic (ROC) curves.

Sparse-View CT Reconstruction Using Curvelet and TV-Based Regularization

The reconstruction from sparse-view projections is one of important problems in computed tomography limited by the availability or feasibility of a large number of projections. Total variation (TV) approaches have been introduced to improve the reconstruction quality by smoothing the variation between neighboring pixels. However, the TV-based methods for images with textures or complex shapes may generate artifacts and cause loss of details. Here, we propose a new regularization model for CT reconstruction by combining regularization methods based on TV and the curvelet transform. Combining curvelet regularizer, which is optimally sparse with better directional sensitivity than wavelet transforms with TV on the other hand will give us a unique regularization model that leads to the improvement of the reconstruction quality. The split-Bregman (augmented Lagrangian) approach has been used as a solver which makes it easy to incorporate multiple regularization terms including the one based on the multiresolution transformation, in our case curvelet transform, into optimization framework. We compare our method with the methods using only TV, wavelet, and curvelet as the regularization terms on the test phantom images. The results show that there are benefits in using the proposed combined curvelet and TV regularizer in the sparse view CT reconstruction.

A CUDA Based Implementation of Locally-and Feature-Adaptive Diffusion Based Image Denoising Algorithm

In this paper we introduce a parallel implementation of locally-and feature-adaptive diffusion based (LFAD) method for image denoising using NVIDIA CUDA framework and graphics processing units (GPUs). LFAD is a novel method for removing additive white Gaussian (AWG) noise in images reported to yield high quality denoised images [1]. It approaches each image region separately and uses different number of nonlinear anisotropic diffusion iterations for each region to attain best peak signal to noise ratio (PSNR). The inverse difference moment (IDM) feature is embedded into a modified diffusion function. As the method has attained highest performance in the class of advanced diffusion based methods and it is competitive with all the state-of-the-art methods, however computationally intensive when executed on the general purpose CPU. To improve the performance, we implemented using the CUDA computational framework. In order to minimize GPU kernel access to the global memory, we use shared memory and the texture memory per multiprocessor. The performance of the GPU implementation of the LFAD has been tested on the standard benchmark images. We demonstrate that with a single NVIDIA Tesla C2050 GPU we can expedite the sequential CPU implementation in most cases from 13 to 20 times.

Channel estimation, equalisation, and evaluation for high-mobility airborne hyperspectral data transmission

In the past few years, unmanned aerial vehicles (UAVs) have become a primary airborne platform for hyperspectral imager for studies on precision agriculture, defence, and the environment. The ‘push-broom’ type of hyperspectral sensors require moving vehicle, and transmission and analysis of hyperspectral data by means of a UAVs high-mobility channel is challenging. While high bandwidth of hyperspectral imaging justify using orthogonal frequency division multiplexing (OFDM) for data transmission, the high speed of UAVs imposes intercarrier interference (ICI) on the transmitted OFDM signal because of the Doppler shift. This study proposes a technique for channel estimation and equalisation in order to compensate the ICI. This technique uses a complete channel matrix estimation in the frequency domain in contrast to conventional methods that only use diagonal elements when recovering the data. In order to evaluate the received data using this technique, a classification framework was designed that took into consideration both spectral and spatial information. In order to verify the robustness of the proposed model, the system was analysed using a Pavia Center hyperspectral dataset, and evaluated against speeds of 50 and 500 m/s. By using this method, improvement in both data transmission and the analysis was achieved.

Sparse-view neutron-photon computed tomography: Object reconstruction and material discrimination

Taking into account the advantages of both neutron- and photon-based systems, we propose combined neutron-photon computed tomography (CT) under a sparse-view setting and demonstrate its performance for 3D object visualization and material discrimination. We use a high-performance regularization method for CT reconstruction by combining regularization based on total variation (TV) and curvelet transform in cone beam geometry. It is coupled with proposed 2D material signatures which is pairs of photon to neutron transmission ratios and neutron transmission values per object space voxels. Classification of materials is performed by association of a voxel signature with library signatures; and per object - by majority of voxels in the object. Representation of object-material pairs, for the model in our experiment, a complex scene with group of high-Z and low-Z materials, attains the reconstruction accuracy of 92.1% and the overall high-Z discrimination accuracy of object representation is 85%, and by about 7.5% higher discrimination accuracy than that with 1D signatures which are ratios of photon to neutron transmissions. With a relative noise level of 10%, the method yields the reconstruction accuracies of 87.2%. The analyses are performed in cone beam configuration, with Monte Carlo modeling of neutron-photon transport for the model of object geometry and material contents.

Evaluation of Audio Denoising Algorithms for Application of Unmanned Aerial Vehicles in Wildlife Monitoring

Unmanned Aerial Vehicles (UAVs) have become popular alternative for wildlife monitoring and border surveillance applications. Elimination of the UAV’s background noise for effective classification of the target audio signal is still a major challenge due to background noise of the vehicles and environments and distances to signal sources. The main goal of this work is to explore acoustic denoising algorithms for effective UAV’s background noise removal. Existing denoising algorithms, such as Adaptive Least Mean Square (LMS), Wavelet Denoising, Time-Frequency Block Thresholding, and Wiener Filter, were implemented and their performance evaluated. LMS and DWT algorithms were implemented on a DSP board and their performance compared using software simulations. Experimental results showed that LMS algorithm’s performance is robust compared to other denoising algorithms. Also, required SNR gain for effective classification of the denosied audio signal is demonstrated.

Compressed sensing magnetic resonance imaging based on shearlet sparsity and nonlocal total variation

Abstract. Compressed sensing (CS) has been utilized for acceleration of data acquisition in magnetic resonance imaging (MRI). MR images can then be reconstructed with an undersampling rate significantly lower than that required by the Nyquist sampling criterion. However, the CS usually produces images with artifacts, especially at high reduction rates. We propose a CS MRI method called shearlet sparsity and nonlocal total variation (SS-NLTV) that exploits SS-NLTV regularization. The shearlet transform is an optimal sparsifying transform with excellent directional sensitivity compared with that by wavelet transform. The NLTV, on the other hand, extends the TV regularizer to a nonlocal variant that can preserve both textures and structures and produce sharper images. We have explored an approach of combining alternating direction method of multipliers (ADMM), splitting variables technique, and adaptive weighting to solve the formulated optimization problem. The proposed SS-NLTV method is evaluated experimentally and compared with the previously reported high-performance methods. Results demonstrate a significant improvement of compressed MR image reconstruction on four medical MRI datasets.

A Trajectory Based Method of Automatic Counting of Cyclist in Traffic Video Data

Due to the growing number of cyclist accidents on urban roads, methods for collecting information on cyclists are of significant importance to the Department of Transportation. The collected information provides insights into solving critical problems related to transportation planning, implementing safety countermeasures, and managing traffic flow efficiently. Intelligent Transportation System (ITS) employs automated tools to collect traffic information from traffic video data. One of the important factors that influence cyclists safety is their counts. In comparison to other road users, such as cars and pedestrians, the automated cyclist data collection is relatively a new research area. In this work, we develop a vision-based method for gathering cyclist count data at intersections and road segments. We implement a robust cyclist detection method based on a combination of classification features. We implement a multi-object tracking method based on the Kernelized Correlation Filters (KCF) in cooperation with the bipartite graph matching algorithm to track multiple cyclists. Then, a trajectory rebuilding method and a trajectory comparison model are applied to refine the accuracy of tracking and counting. The proposed method is the first cyclist counting method, that has the ability to count cyclists under different movement patterns. The trajectory data obtained can be further utilized for cyclist behavioral modeling and safety analysis.

A Deep Belief Network for Classifying Remotely-Sensed Hyperspectral Data

Improving the classification accuracy of remotely sensed data is of paramount interest for science and defense applications. In this paper, we investigate deep learning architectures (DLAs), whose popularity has grown recently due to the discovery of efficient algorithms to train them, one of which, unsupervised pre-training, seeks to initialize the learned model in a way that greatly encourages efficient supervised learning. We propose a structure for a DLA, the deep belief network (DBN), suitable for the classification of remotely-sensed hyperspectral data. To arrive at this structure, we first study the role of the DBN’s width and the duration of pre-training in the learning of features used for the multiclass discrimination of spectral data. We then study the effect of exploiting joint spectral-spatial information. The support vector machine (SVM) is used as a baseline to determine that the proposed method is feasible, offering consistently high classification accuracies in comparison.

Bicycle Detection Using HOG, HSC and MLBP

Due to the growing number of bicycles on roads, safety of bicyclists is drawing the increasing attention of transportation departments. Intelligent Transportation Systems (ITS) use automated tools for processing and analysis of traffic video data to plan and implement safety measures. One of important factors that influence the planning and safety countermeasures for bicyclists is the bicycle count. In this paper, we develop a bicycle detection method that can be used in a bicycle counting system. We strive to improve the efficiency of detection by looking for classification features that deliver more versatile information to automatic classifiers. We explore a combination of Histograms of Oriented Gradients (HOG), Histogram of Shearlet Coefficients (HSC) and Multi-scale Local Binary Pattern (MLBP) to improve detection and count of bicycles in video data. It is shown that the combination of the above features secures a higher detection accuracy.