M. Lucena

A scheme of colour image retrieval from databases

Abstract This paper presents a scheme of image retrieval from a database using queries prompted by the colour and the shape of the objects present in different scenes. Of the whole scheme of image retrieval, we will focus attention on the modules that allow feature extraction of the component objects from the scenes and the matching of the objects among the different images. The defined scheme enables the indexing of images by measuring the similarity between the integral objects.

Human Action Recognition Using Optical Flow Accumulated Local Histograms

This paper addresses the human action recognition task from optical flow. We develop a non-parametric motion model using only the image region surrounding the actor making the action. For every two consecutive frames, a local motion descriptor is calculated from the optical flow orientation histograms collected from overlapping regions inside the bounding box of the actor. An action descriptor is built by weighting and aggregating the estimated histograms along the temporal axis. We obtain a promising trade-off between complexity and performance compared with state-of-the-art approaches. Experimental results show that the proposed method equals or improves on the performance of state-of-the-art approaches using these databases.

Human action recognition based on aggregated local motion estimates

This paper addresses the human action recognition task from optical flow. This task is in itself an interesting problem, given the lack of accuracy and noisy characteristics of the optical flow estimation. Optical flow is one of the most popular descriptors characterizing motion, but due to its instability is usually used in combination with parametric models. In this work, we develop a non-parametric motion model using only the image region surrounding the actor making the action. To be precise, for every two consecutive frames, a local motion descriptor is calculated from the optical flow orientation histograms collected inside the actor’s bounding box. An action descriptor is built by weighting and aggregating the estimated histograms along the temporal axis. The proposed approach obtains a promising trade-off between complexity and performance compared with state-of-the-art approaches. The action recognition can also be done in real time by accumulating evidence from each new incoming image. Experiments on two well-known video sequence databases are carried out in order to evaluate the behavior of the proposal.

Evaluation of three optical flow-based observation models for tracking

We study the use of optical flow as a characteristic for tracking. We analyze the behavior of three flow-based observation models for particle filter algorithms, and compare the results with those obtained using a well-known, gradient-based, observation model. Although in theory, optical flow could be used directly to displace an object model, in practice, flow estimation techniques lack the necessary precision. In view of the fact that probabilistic tracking algorithms enable imprecise or incomplete information to be handled naturally, these models have been used as a natural way of incorporating flow information into the tracking.

Tracking people in video sequences using multiple models

This paper presents a multiple model real-time tracking technique for video sequences, based on the mean-shift algorithm. The proposed approach incorporates spatial information from several connected regions into the histogram-based representation model of the target, and enables multiple models to be used to represent the same object. The use of several regions to capture the color spatial information into a single combined model, allow us to increase the object tracking efficiency. By using multiple models, we can make the tracking scheme more robust in order to work with sequences with illumination and pose changes. We define a model selection function that takes into account both the similarity of the model with the information present in the image, and the target dynamics. In the tracking experiments presented, our method successfully coped with lighting changes, occlusion, and clutter.

Decision support system for classifying archaeological pottery profiles based on Mathematical Morphology

We present a decision support system to help archaeologists in classifying wheel-made pottery pieces by its profile. A novel shape characterization method, using Mathematical Morphology, is introduced for this purpose. Each profile is represented as a vector, obtained by sampling the so called morphological curves (erosion, dilation, opening and closing), and Euclidean Distance is used as a similarity measure. We show results of our method applied to a profile database of Iberian Pottery from the upper valley of Guadalquivir River (Spain).

Optical flow-based observation models for particle filter tracking

This paper presents three observation models suitable for particle filter tracking, based on the optical flow of the sequence. Modern optical flow computation techniques can obtain in real time very accurate estimates, so we can use it as a source of evidence for higher level image processing. Our image motion-based models are based, respectively, on: a previously computed optical flow field, the image brightness constraint, and similarity measures. They take into account not only the consistency of the measured optical flow with the motion predicted by the model, but also the presence of optical flow discontinuities on the object boundary. Experimental results show that the resulting trackers are comparable to other, state-of-the-art tracking methods. While the model based on similarity measures provides better performance, the optical flow-field-based model is a suitable option when the flow field is available.

Morphometric Analysis Applied to the Archaeological Pottery of the Valley of Guadalquivir

Ceramics are one of the most documented materials in the archaeological interventions. The documentation and the analysis of the pottery shapes allow the knowledge of the chronology and the functionality of the settlement where they have been found.

Learning action descriptors for recognition

This paper evaluates different Restricted Boltzmann Machines models in unsupervised, semi-supervised and supervised frameworks using information from human actions. After feeding these multilayer models with low level features, we infer high-level discriminating features that highly improve the classification performance. This approach eliminates the difficult process of selecting good mid-level feature descriptors, changing the feature selection and extraction process by a learning stage. Two main contributions are presented. First, a new sequence-descriptor from accumulated histograms of optical flow (aHOF) is presented. Second, comparative results using unsupervised, supervised and semi-supervised classification experiments are shown. The results show that the RBM architectures provide very good results in our classification task and present very good properties for semi-supervised learning.

Using optical flow as evidence for probabilistic tracking

In this paper, we presen t an observation model based on the Lucas and Kanade algorithm for computing optical flow, to trac k objects using particle filter algorithms. Although optical flow information enables us to know the displacement of objects present in a scene, it cannot be used directly to displace an object model since flow calculation techniques lack the necessary precision. In view of the fact that probabilistic tracking algorithms enable imprecise or incomplete information to be handled naturally, this model has been used as a natural means of incorporating flow information into the trac king.