Extracting image features for classification by two-tier genetic programming

Abstract

Image classification is a complex but important task especially in the areas of machine vision and image analysis such as remote sensing and face recognition. One of the challenges in image classification is finding an optimal set of features for a particular task because the choice of features has direct impact on the classification performance. However the goodness of a feature is highly problem dependent and often domain knowledge is required. To address these issues we introduce a Genetic Programming (GP) based image classification method, Two-Tier GP, which directly operates on raw pixels rather than features. The first tier in a classifier is for automatically defining features based on raw image input, while the second tier makes decision. Compared to conventional feature based image classification methods, Two-Tier GP achieved better accuracies on a range of different tasks. Furthermore by using the features defined by the first tier of these Two-Tier GP classifiers, conventional classification methods obtained higher accuracies than classifying on manually designed features. Analysis on evolved Two-Tier image classifiers shows that there are genuine features captured in the programs and the mechanism of achieving high accuracy can be revealed. The Two-Tier GP method has clear advantages in image classification, such as high accuracy, good interpretability and the removal of explicit feature extraction process.

Automatically Evolving Rotation-invariant Texture Image Descriptors by Genetic Programming

“…Hence, their two-tier GP (2T-GP) approach is quite similar to the 3T-GP approach but with more terminals and functions added [30], [43]. The 2T-GP has been tested using datasets for different applications and it was shown to outperform the competitor methods [43].…”

“…The 2T-GP has been tested using datasets for different applications and it was shown to outperform the competitor methods [43]. Moreover, using the features extracted by 2T-GP improved the performance of the different classifiers compared to the use of domain-specific hand-crafted features [30].…”

Subset Selection for Landmark Modern and Historic Images

“…al. [19] introduced a Two-Tier Genetic Programming (GP) based image classification method which works on raw pixels rather than high-level features.…”

Evolutionary Multiobjective Image Feature Extraction in the Presence of Noise

“…These features are not necessarily similar to the human experts with domain knowledge, but can still perform similarly well. Generally, features extracted using evolutionary algorithms are more efficient to compute and perform reasonably well compared to manually constructed features, see [30] and [31] for example and references cited therein.…”