Víctor González-Castro

Class distribution estimation based on the Hellinger distance

Class distribution estimation (quantification) plays an important role in many practical classification problems. Firstly, it is important in order to adapt the classifier to the operational conditions when they differ from those assumed in learning. Additionally, there are some real domains where the quantification task is itself valuable due to the high variability of the class prior probabilities. Our novel quantification approach for two-class problems is based on distributional divergence measures. The mismatch between the test data distribution and validation distributions generated in a fully controlled way is measured by the Hellinger distance in order to estimate the prior probability that minimizes this divergence. Experimental results on several binary classification problems show the benefits of this approach when compared to such approaches as counting the predicted class labels and other methods based on the classifier confusion matrix or on posterior probability estimations. We also illustrate these techniques as well as their robustness against the base classifier performance (a neural network) with a boar semen quality control setting. Empirical results show that the quantification can be conducted with a mean absolute error lower than 0.008, which seems very promising in this field.

Texture and moments-based classification of the acrosome integrity of boar spermatozoa images

The automated assessment of the sperm quality is an important challenge in the veterinary field. In this paper, we explore how to describe the acrosomes of boar spermatozoa using image analysis so that they can be automatically categorized as intact or damaged. Our proposal aims at characterizing the acrosomes by means of texture features. The texture is described using first order statistics and features derived from the co-occurrence matrix of the image, both computed from the original image and from the coefficients yielded by the Discrete Wavelet Transform. Texture descriptors are evaluated and compared with moments-based descriptors in terms of the classification accuracy they provide. Experimental results with a Multilayer Perceptron and the k-Nearest Neighbours classifiers show that texture descriptors outperform moment-based descriptors, reaching an accuracy of 94.93%, which makes this approach very attractive for the veterinarian community.

Color Adaptive Neighborhood Mathematical Morphology and its application to pixel-level classification

In this paper spatially adaptive Mathematical Morphology (MM) is studied for color images. More precisely, the General Adaptive Neighborhood Image Processing (GANIP) approach is generalized to color images. The basic principle is to define a set of locally Color Adaptive Neighborhoods (CAN), one for each point of the image, and to use them as adaptive structuring elements (ASE) for morphological operations. These operators have been applied to images in different color spaces and compared with other kinds of ASEs extended to color images. Results show that the proposed method is more respectful with the borders of the objects, as well as with the color transitions within the image. Finally, the proposed adaptive morphological operators are applied to the classification of color texture images.

Local Oriented Statistics Information Booster (LOSIB) for Texture Classification

Local oriented statistical information booster (LOSIB) is a descriptor enhancer based on the extraction of the gray level differences along several orientations. Specifically, the mean of the differences along particular orientations is considered. In this paper we have carried out some experiments using several classical texture descriptors to show that classification results are better when they are combined with LOSIB, than without it. Both parametric and non-parametric classifiers, Support Vector Machine and k-Nearest Neighbourhoods respectively, were applied to assess this new method. Furthermore, two different texture dataset were evaluated: KTH-Tips-2a and Brodatz32 to prove the robustness of LOSIB. Global descriptors such as WCF4 (Wavelet Co-occurrence Features), that extracts Haralick features from the Wavelet Transform, have been combined with LOSIB obtaining an improvement of 16.94% on KTH and 7.55% on Brodatz when classifying with SVM. Moreover, LOSIB was used together with state-of-the-art local descriptors such as LBP (Local Binary Pattern) and several of its recent variants. Combined with CLBP (Complete LBP), the LOSIB booster results were improved in 5.80% on KTH-Tips 2a and 7.09% on the Brodatz dataset. For all the tested descriptors, we have observed that a higher performance has been achieved, with the two classifiers on both datasets, when using some LOSIB settings.

Curvelet-based texture description to classify intact and damaged boar spermatozoa

The assessment of boar sperm head images according to their acrosome status is a very important task in the veterinary field. Unfortunately it can only be performed manually, which is slow, non-objective and expensive. It is important to provide companies an automatic and reliable method to perform this task. In this paper a new method which uses texture descriptors based on the Curvelet Transform is proposed. Its performance has been compared with other texture descriptors based on the Wavelet transform, and also with moments based descriptors, as they seem to be successful for this problem. Texture descriptors performed better, and curvelet-based ones achieved the best hit rate (97%) and area under the ROC curve (0.99).

Boar spermatozoa classification using longitudinal and transversal profiles (LTP) descriptor in digital images

A new textural descriptor, named Longitudinal and Transversal Profiles (LTP), has been proposed. This descriptor was used to classify 376 images of dead spermatozoa heads and 472 images of alive ones. The result obtained with this descriptor has been compared with the Pattern spectrum, Flusser, Hu, and a descriptor based on statistical values of the histogram. The features vectors computed have been classified using a back-propagation Neural Network and the kNN (k Nearest Neighbours) algorithm. Classification error obtained with LTP was 30.58% outperforming the other descriptors. The area under the ROC curve (AUC) has also been calculated confirming that the performance of the proposed descriptor is better that of the other texture descriptors.

Estimating class proportions in boar semen analysis using the hellinger distance

Advances in image analysis make possible the automatic semen analysis in the veterinary practice. The proportion of sperm cells with damaged/intact acrosome, a major aspect in this assessment, depends strongly on several factors, including animal diversity and manipulation/ conservation conditions. For this reason, the class proportions have to be quantified for every future (test) semen sample. In this work, we evaluate quantification approaches based on the confusion matrix, the posterior probability estimates and a novel proposal based on the Hellinger distance. Our information theoretic-based approach to estimate the class proportions measures the similarity between several artificially generated calibration distributions and the test one at different stages: the data distributions and the classifier output distributions. Experimental results show that quantification can be conducted with a Mean Absolute Error below 0.02, what seems promising in this field.

Application of the ordered logit model to optimising Frangi filter parameters for segmentation of perivascular spaces

Segmentation of perivascular spaces (PVS) from brain magnetic resonance images (MRI) is important for understanding the brains lymphatic system and its relationship with neurological diseases. The Frangi filter might be a valuable tool for this purpose. However, its parameters need to be adjusted in response to the variability in the scanners parameters and study protocols. Knowing the neuroradiological ratings of the PVS, we used the ordered logit model to optimise Frangi filter parameters. The PVS volume obtained significantly and strongly correlated with neuroradiological assessments (Spearmans ρ=0.75, p < 0.001), suggesting that the ordered logit model could be a good alternative to conventional optimisation frameworks for segmenting PVS on MRI.

Compass radius estimation for improved image classification using Edge-SIFT

The combination of SIFT descriptors with other features usually improves image classification, like Edge-SIFT, which extracts keypoints from an edge image obtained after applying the compass operator to a colour image. We evaluate for the first time, how the use of different radii in the compass operator affects the classification performance. We demonstrate that the value proposed in the literature, radius=4.00, is not the optimum from an image classification point of view. We also put in evidence that in ideal conditions, choosing an appropriate radius for each image yields accuracy values even higher than 95%. Finally, we propose a new method to estimate the best radius for the compass operator in each dataset. Using a training subset selected on the basis of a minimum dispersion criterion of edges density, we construct a richer dictionary for each dataset in our Bag of Words pipeline. From that dictionary it is selected a radius for the whole dataset that yields higher accuracy than using the value proposed in the literature. Using this method, we obtained improvements in the accuracy up to 24.4% in Soccer, 6.77% in COIL-RWTH-2, 4.46% in Birds, 3.82% in ImageNet_Dogs, 2.75% in ImageNet_Birds, 2.02% in Flowers and 1.75% in Caltech101 datasets. It was demonstrated that compass radius in Edge-SIFT affects to classification.The classification performance of different radii was evaluated on eight datasets.It is shown that selecting a radius for each image results in better classification.A method to automatically estimate a better radius for each dataset is proposed.The estimated radius guarantees better results than the state-of-the-art.

aZIBO: A New Descriptor Based in Shape Moments and Rotational Invariant Features

In this work, a descriptor called a ZIBO (absolute Zernike moments with Invariant Boundary Orientation) that describes the shape of objects using the module of Zernike moments and the edge features obtained from an almost rotational invariant version of the Edge Gradient Co-occurrence Matrix (EGCM) is proposed. The two descriptors obtained, the Zernike module as global descriptor and the new version of EGCM as local one, are used to characterize images from three different datasets, Kimia99, MPEG2 and MPEG7. Later on, the concatenation of both local and global descriptors was evaluated using kNN with City block and Chi-square distance metrics. Also, the descriptors are assessed separately with a weight-based method, being the results obtained compared with the ones reached by the baseline method, ZMEG (Zernike Moment Edge Gradient). Using MPEG7, which is the most challenging dataset, and the weight-based classifier, this proposal obtained a success rate of 78.29%, outperforming the 75.86% achieved by ZMEG method. With the MPEG2 dataset, results were even better with an 81.00% of success rate against 77.25% of ZMEG.