Golrokh Mirzaei

The application of Evolutionary Neural Network for bat echolocation calls recognition

An Evolutionary Neural Network (ENN) is developed to identify bats by their vocalization characteristics. This is in an effort to identify local bat species as a large number of bat fatalities near wind turbines have been reported. ENN is based on the Genetic Algorithm, which can be used for optimization of the weight selection of the neural network. We then compare ENN with different classification techniques. In the scope of bat call classification, ENN is a new technique that can be effectively used as a bat-call classifier. This research will help in developing mitigation techniques for reducing bat fatalities. The ENN algorithm is developed in MATLAB.

Avian detection & tracking algorithm using infrared imaging

This paper presents a method for target detection and tracking of IR images in the application of avian surveillance. As there are many reports of avian mortality due to collision with turbine blades, the detection and tracking of birds at turbine sites is an important issue. In this work, three different background subtraction techniques are first applied to detect moving objects. Otsu thresholding method is then extended by incorporating an adaptive variable based on the mean of each frame and certain constant value. Filtering using morphological operations is applied. Results of three different techniques are then compared. Selected technique (RA) followed by thresholding and filtering is then used for tracking and information extraction. Results show that proposed method provides the needed accuracy for IR imagery. This method can be effectively used in different applications of IR imaging.

Implementation of ant clustering algorithm for IR imagery in wind turbine applications

Interaction of avian with turbines has become an important public policy issue, so identification and quantification of avian at turbine sites is crucial.

A novel feature extraction algorithm for classification of bird flight calls

Acoustic monitoring of birds in the vicinity of wind turbines is becoming an important public policy issue. Acoustic monitoring involves preprocessing, feature extraction and classification. A novel Spectrogram-based Image Frequency Statistics (SIFS) feature extraction algorithm has been developed. Features extracted from proposed algorithms were then combined with various classification algorithms such as k-NN, Multilayer Perceptron (MLP) and Hidden Markov Models (HMM) and Evolutionary Neural Network (ENN). SIFS and MMS algorithms, combined with ENN, provided the most accurate results. Proposed algorithms were tested with real data collected during spring migration around Lake Erie in Ohio.

Birds/bats movement tracking with IR camera for wind farm applications

Nocturnally migratory birds and bats are at higher risk of colliding with wind turbines. It is important to gather scientific data in an area which have potential of wind farm development. An IR camera recording and its analysis can provide necessary information to wildlife biologists involve with interaction of birds/bats with wind turbines. An efficient IR video processing algorithm has been developed. The proposed algorithm consists of background and consecutive frame subtraction, frame selection, 3-D region labeling and breakpoint recovery. It is then used to process spring 2011 bird migration data that has been collected in Ottawa National Wildlife Refuge in Ohio. Results from this study will be useful for wildlife biologists to make intelligent decision for siting of wind turbines. It will also help policy makers to develop an appropriate public policy for wind farm development in an area with extensive avian activity.

The BIO-acoustic feature extraction and classification of bat echolocation calls

There are reports that large number of bat fatalities occur near wind turbines. Acoustic characteristics can be employed for bat call recognition to better understand the effects of turbines on different bat species. Acoustic features of bat echolocation calls are extracted based on three different techniques: Short Time Fourier Transform (STFT), Mel Frequency Cepstrum Coefficient (MFCC) and Discrete Wavelet Transform (DWT). These features are fed into an Evolutionary Neural Network (ENN) for their classification at the species level using acoustic features. Results from different feature extraction techniques are compared based on classification accuracy. The technique can identify bats and will contribute towards developing mitigation procedures for reducing bat fatalities.

Acoustic monitoring techniques for avian detection and classification

There are many reports of bird and bat mortality in vicinity of wind turbines [1]. It is important to quantify numbers and species of birds and bats in a given area which is targeted for wind farm development. It is also necessary to assess the behavior of birds and bats in wind farm areas. Acoustic monitoring techniques have been developed in this work for monitoring of birds and bats. Spectrogram-based Image Frequency Statistics (SIFS) is used for feature extraction and Evolutionary Neural Network (ENN) is used for classification purposes. Data was collected near Lake Erie in Ohio during 2011 spring and fall migration periods. Data analysis was performed in accordance to needs of wildlife biologists.

Radar-based monitoring system for nocturnal assessment

A radar monitoring system is implemented to evaluate the birds activity and their quantification in the western basin of lake Erie in Ohio. This location is a habitat stopover for migratory birds and also it is a putative area for constructing wind farm. The radar monitoring is performed in three steps consist of data acquisition, blip detections, and target tracking. Blip detection is performed in radR platform and tracking is performed using developed particle filter in MATLAB. The radar monitoring system is used to process data collected in migration period of fall and spring 2011.

Data fusion of IR and marine radar data

Avian monitoring system is developed based on data fusion of thermal/Infrared camera (IR) and marine radar. First data were processed separately using video/image processing and radar signal processing techniques and features of the targets are obtained by each sensor. Data fusion of radar and IR then is implemented to achieve feature vectors of the target. IR camera provides the coordinate information of the targets as well as some features such as flights straightness index, direction, and targets heat, while the radar provides altitude information (z-coordinates) which is not provided by IR. So the data fusion of the IR and radar provide more detail and reliable information possible of the avian targets and their activity. Data was collected near Lake Erie in Ohio during 2011 spring and fall migration periods. Data analysis was performed in accordance to needs of wildlife biologists.

Evolutionary Neural Network parallelization with multicore systems on chip

Evolutionary Neural Network (ENN) has attracted great attention among the researchers in recent years because of its effectiveness at function optimization and, its efficiency in searching large and complex spaces to find nearly global optima. In this work, Parallel Evolutionary Neural Network algorithm is proposed and implemented on Multi-core system on chip. The algorithm is parallelized, partitioned, mapped, and scheduled on multicore. The algorithm is also implemented on single core for comparison. The parallel ENN is developed in C# using .Net framework 4.0. The .Net framework offers comprehensive and flexible threads APIs that allow the efficient implementation of multithreaded applications.