R. da S. Torres

A graph-based approach for multiscale shape analysis

Abstract This paper presents two shape descriptors, multiscale fractal dimension and contour saliences, using a graph-based approach— the image foresting transform. It introduces a robust approach to locate contour saliences from the relation between contour and skeleton. The contour salience descriptor consists of a vector, with salience location and value along the contour, and a matching algorithm. We compare both descriptors with fractal dimension, Fourier descriptors, moment invariants, Curvature Scale Space and Beam Angle Statistics regarding to their invariance to object characteristics that belong to a same class (compact-ability) and to their ability to separate objects of distinct classes (separability).

Contour salience descriptors for effective image retrieval and analysis

Abstract This work exploits the resemblance between content-based image retrieval and image analysis with respect to the design of image descriptors and their effectiveness. In this context, two shape descriptors are proposed: contour saliences and segment saliences . Contour saliences revisits its original definition, where the location of concave points was a problem, and provides a robust approach to incorporate concave saliences. Segment saliences introduces salience values for contour segments, making it possible to use an optimal matching algorithm as distance function. The proposed descriptors are compared with convex contour saliences, curvature scale space , and beam angle statistics using a fish database with 11,000 images organized in 1100 distinct classes. The results indicate segment saliences as the most effective descriptor for this particular application and confirm the improvement of the contour salience descriptor in comparison with convex contour saliences.

Relevance feedback based on genetic programming for image retrieval

This paper presents two content-based image retrieval frameworks with relevance feedback based on genetic programming. The first framework exploits only the user indication of relevant images. The second one considers not only the relevant but also the images indicated as non-relevant. Several experiments were conducted to validate the proposed frameworks. These experiments employed three different image databases and color, shape, and texture descriptors to represent the content of database images. The proposed frameworks were compared, and outperformed six other relevance feedback methods regarding their effectiveness and efficiency in image retrieval tasks.

A relevance feedback method based on genetic programming for classification of remote sensing images

This paper presents an interactive technique for remote sensing image classification. In our proposal, users are able to interact with the classification system, indicating regions of interest (and those which are not). This feedback information is employed by a genetic programming approach to learning user preferences and combining image region descriptors that encode spectral and texture properties. Experiments demonstrate that the proposed method is effective for image classification tasks and outperforms the traditional MaxVer method.

Multiscale Classification of Remote Sensing Images

A huge effort has been applied in image classification to create high-quality thematic maps and to establish precise inventories about land cover use. The peculiarities of remote sensing images (RSIs) combined with the traditional image classification challenges made RSI classification a hard task. Our aim is to propose a kind of boost-classifier adapted to multiscale segmentation. We use the paradigm of boosting, whose principle is to combine weak classifiers to build an efficient global one. Each weak classifier is trained for one level of the segmentation and one region descriptor. We have proposed and tested weak classifiers based on linear support vector machines (SVM) and region distances provided by descriptors. The experiments were performed on a large image of coffee plantations. We have shown in this paper that our approach based on boosting can detect the scale and set of features best suited to a particular training set. We have also shown that hierarchical multiscale analysis is able to reduce training time and to produce a stronger classifier. We compare the proposed methods with a baseline based on SVM with radial basis function kernel. The results show that the proposed methods outperform the baseline.

Shape feature extraction and description based on tensor scale

Tensor scale is a morphometric parameter that unifies the representation of local structure thickness, orientation, and anisotropy, which can be used in several computer vision and image processing tasks. In this article, we exploit this concept for binary images and propose a shape salience detector and a shape descriptor-Tensor Scale Descriptor with Influence Zones. It also introduces a robust method to compute tensor scale, using a graph-based approach-the Image Foresting Transform. Experimental results are provided, showing the effectiveness of the proposed methods, when compared to other relevant methods, such as Beam Angle Statistics and Contour Salience Descriptor, with regard to their use in content-based image retrieval tasks.

Rotation-Invariant and Scale-Invariant Steerable Pyramid Decomposition for Texture Image Retrieval

This paper proposes a new rotation-invariant and scale-invariant representation for texture image retrieval based on steerable pyramid decomposition. By calculating the mean and standard deviation of decomposed image subbands, the texture feature vectors are extracted. To obtain rotation or scale invariance, the feature elements are aligned by considering either the dominant orientation or dominant scale of the input textures. Experiments were conducted on the Brodatz database aiming to compare our approach to the conventional steerable pyramid decomposition, and a proposal for texture characterization based on Gabor wavelets with regard to their retrieval effectiveness. Results demonstrate the superiority of the proposed method in rotated and scaled image datasets.In highly tessellated models, triangles are very small compared to the entire object, representing at the same time its macro- and mesostructures. The main idea in this work is to use a visualization algorithm that is adequate to mesostructure but applied to the whole object. Tessellated models are converted into geometry textures, a geometric representation for surfaces based on height maps. In rendering time, the fine-scale details are reconstructed with LOD speed-up while preserving original quality.

Color Descriptors for Web Image Retrieval: A Comparative Study

This paper presents a comparative study of color descriptors for content-based image retrieval on the Web. Several image descriptors were compared theoretically and the most relevant ones were implemented and tested in two different databases. The main goal was to find out the best descriptors for Web image retrieval. Descriptors are compared according to the extraction and distance functions complexities, the compactness of feature vectors, and the ability to retrieve relevant images.

TSD: A Shape Descriptor Based on a Distribution of Tensor Scale Local Orientation

We present tensor scale descriptor (TSD) — a shape descriptor for content-based image retrieval, registration, and analysis. TSD exploits the notion of local structure thickness, orientation, and anisotropy as represented by the largest ellipse centered at each image pixel and within the same homogeneous region. The proposed method uses the normalized histogram of the local orientation (the angle of the ellipse) at regions of high anisotropy and thickness within a certain interval. It is shown that TSD is invariant to rotation and to some reasonable level of scale changes. Experimental results with a fish database are presented to illustrate and validate the method.

Automatic identification of fruit flies (Diptera

Description and learning techniques for fruit fly identification.Image analysis techniques for classifying wings and aculeus.A novel multimodal classification approach for fruit fly identification based on wing and aculeus.Good results in the identification of fruit flies based on wings, aculei, and their combination. Fruit flies are pests of major economic importance in agriculture. Among these pests it is possible to highlight some species of genus Anastrepha, which attack a wide range of fruits, and are widely distributed in the American tropics and subtropics. Researchers seek to identify fruit flies in order to implement management and control programs as well as quarantine restrictions. However, fruit fly identification is manually performed by scarce specialists through analysis of morphological features of the mesonotum, wing, and aculeus. Our objective is to find solid knowledge that can serve as a basis for the development of a sounding automatic identification system of the Anastrepha fraterculus group, which is of high economic importance in Brazil. Wing and aculeus images datasets from three specimens have been used in this work. The experiments using a classifier multimodal fusion approach shows promising effectiveness results for identification of these fruit flies, with more than 98% classification accuracy, a remarkable result for this difficult problem.