Comparison of GENIE and conventional supervised classifiers for multispectral image feature extraction

  title={Comparison of GENIE and conventional supervised classifiers for multispectral image feature extraction},
  author={Neal R. Harvey and James Theiler and Steven P. Brumby and Simon Perkins and John J. Szymanski and Jeffrey J. Bloch and Reid B. Porter and Mark Galassi and A. Cody Young},
  journal={IEEE Trans. Geosci. Remote. Sens.},
The authors have developed an automated feature detection/classification system, called GENetic Imagery Exploitation (GENIE), which has been designed to generate image processing pipelines for a variety of feature detection/classification tasks. GENIE is a hybrid evolutionary algorithm that addresses the general problem of finding features of interest in multispectral remotely-sensed images. The authors describe their system in detail together with experiments involving comparisons of GENIE… 

Figures and Tables from this paper

Automated simultaneous multiple feature classification of MTI data
The improvements made to the GENIE software to allow multiple-feature classification are described and the application of this system to the automatic simultaneous classification of multiple features from MTI image data is described.
Multimodal approach to feature extraction for image and signal learning problems
We present ZEUS, an algorithm for extracting features from images and time series signals. ZEIS is designed to solve a variety of machine learning problems including time series forecasting, signal
Evolving spatio-spectral feature extraction algorithms for hyperspectral imagery
Hyperspectral imagery data sets present an interesting challenge to feature extraction algorithm developers. Beyond the immediate problem of dealing with the sheer amount of spectral information per
Unsupervised land cover classification in multispectral imagery with sparse representations on learned dictionaries
The results suggest that neuroscience-based models are a promising approach to practical pattern recognition problems in remote sensing, even for datasets using spectral bands not found in natural visual systems.
Evolving feature extraction algorithms for hyperspectral and fused imagery
A genetic programming software system, called GENIE, is described, which augments the human scientist/analyst by evolving customized spatio-spectral feature extraction pipelines from training data provided via an intuitive, point-and-click interface.
Robust support vector method for hyperspectral data classification and knowledge discovery
Support vector machines yield better outcomes than neural networks regarding accuracy, simplicity, and robustness, and training neural and neurofuzzy models is unfeasible when working with high-dimensional input spaces and great amounts of training data.
Discovering Rules with Genetic Algorithms to Classify Urban Remotely Sensed Data
The possibilities of using genetic programming for the automatic extraction of classification rules from urban remotely sensed data and the quality of the results shows the effectiveness of the proposed genetic classifier in the object-oriented, knowledge-based approach.
Feature extraction from multiple data sources using genetic programming
Feature extraction from imagery is an important and long-standing problem in remote sensing. In this paper, we report on work using genetic programming to perform feature extraction simultaneously
Land cover classification in multispectral satellite imagery using sparse approximations on learned dictionaries
Results from an ongoing effort to extend machine vision methodologies to the environmental sciences, using state-of-theart adaptive signal processing, combined with compressive sensing and machine learning techniques are presented.
Land cover classification in multispectral imagery using clustering of sparse approximations over learned feature dictionaries
This work uses a Hebbian learning rule to build spectral-textural dictionaries that are tailored for classification and explores a quantitative metric to evaluate the spectral properties of the clusters in order to potentially aid in assigning land cover categories to the cluster labels.


GENIE: a hybrid genetic algorithm for feature classification in multispectral images
We consider the problem of pixel-by-pixel classification of a multi- spectral image using supervised learning. Conventional spuervised classification techniques such as maximum likelihood
Image classification using spectral and spatial information based on MRF models
A new criterion for classifying multispectral remote sensing images or textured images by using spectral and spatial information is proposed and a stepwise classification algorithm is derived.
Unsupervised retraining of a maximum likelihood classifier for the analysis of multitemporal remote sensing images
The proposed technique allows the classifier's parameters to be updated in a totally unsupervised way on the basis of the distribution of a new image to be classified to provide a high accuracy for the new image even when the corresponding training set is not available.
Genetic algorithm for combining new and existing image processing tools for multispectral imagery
This work describes the implementation and performance of a genetic algorithm (GA) which evolves and combines image processing tools for multispectral imagery (MSI) datasets, and describes the chromosomal representation of candidate image processing algorithms, and discusses the set of image operators.
Investigation of image feature extraction by a genetic algorithm
The implementation and performance of a genetic algorithm which generates image feature extraction algorithms for remote sensing applications and the basis set of primitive image operators and chromosomal representation of a complete algorithm are described.
Parallel evolution of image processing tools for multispectral imagery
A parallel, hybrid evolutionary-algorithm-based system, which optimizes image processing tools for feature-finding tasks in multi-spectral imagery (MSI) data sets and is capable of combining suitably-registered data from different sensors is described.
Classification of multispectral data by joint supervised-unsupervised learning
It is shown that by adding the unlabeled samples to the classifier design process, better estimates for the discriminant functions can be obtained and the peaking phenomenon that is observed in the performance versus dimensionality curves, can be mitigated.
Statistics enhancement in hyperspectral data analysis using spectral-spatial labeling, the EM algorithm, and the leave-one-out covariance estimator
Hyperspectral data potentially contain more information than multispectral data because of higher dimensionality. Information extraction algorithm performance is strongly related to the quantitative
Pixel classification using variable string genetic algorithms with chromosome differentiation
The concept of chromosome differentiation, commonly witnessed in nature as male and female sexes, is incorporated in genetic algorithms with variable length strings for designing a nonparametric
Contiguity-enhanced k-means clustering algorithm for unsupervised multispectral image segmentation
This work suggests a simple approach - a variant of the standard k-means algorithm - which uses both spatial and spectral properties of the image, and finds that the spatial contiguity and spectral compactness properties are nearly 'orthogonal', which means that it can make considerable improvements in the one with minimal loss in the other.