International audienceIn this paper, we propose a novel phased-array track before detect (TBD) filter for tracking multiple distributed (extended) targets from impulsive observations. Since the targets are angularly spread, we track the centroid Direction Of Arrival (DOA) of the target-generated (or backscattered) signal. The main challenge stems from the random target signals that, conditional to their respective states, constitute non-deterministic contributions to the system observation. The novelty of our approach is twofold. First, we develop a Cardinalized Probability Hypothesis Density (CPHD) filter for tracking multiple targets with non-deterministic contributions, more specifically, Spherically Invariant RandomVector (SIRV) processes. This is achieved by analytically integrating the SIRV and angularly distributed target signals in the update step. Thus, ensuring a more efficient implementation than existing solutions, that generally consider augmenting the target state with the target signal. Secondly, we provide an improved auxiliary particle CPHD filter and clustering methodology. The auxiliary step is carried out for persistent particles, while for newly birthed particles an adaptive importance distribution is given. This is in contrast with existing solutions that only consider the auxiliary step for birthed particles. Simulated data results showcase the improved performance of the proposed filter. Results on real sonar phased-array data are presented for underwater 3D image reconstruction applications

Automatic extraction of buildings in urban and residential scenes has become a subject of growing interest in the domain of photogrammetry and remote sensing, particularly since the mid-1990s. Active contour model, colloquially known as snake model, has been studied to extract buildings from aerial and satellite imagery. However, this task is still very challenging due to the complexity of building size, shape, and its surrounding environment. This complexity leads to a major obstacle for carrying out a reliable large-scale building extraction, since the involved prior information and assumptions on building such as shape, size, and color cannot be generalized over large areas. This paper presents an efficient snake model to overcome such a challenge, called Super-Resolution-based Snake Model (SRSM). The SRSM operates on high-resolution Light Detection and Ranging (LiDAR)-based elevation images—called z-images—generated by a super-resolution process applied to LiDAR data. The involved balloon force model is also improved to shrink or inflate adaptively, instead of inflating continuously. This method is applicable for a large scale such as city scale and even larger, while having a high level of automation and not requiring any prior knowledge nor training data from the urban scenes (hence unsupervised). It achieves high overall accuracy when tested on various datasets. For instance, the proposed SRSM yields an average area-based Quality of 86.57% and object-based Quality of 81.60% on the ISPRS Vaihingen benchmark datasets. Compared to other methods using this benchmark dataset, this level of accuracy is highly desirable even for a supervised method. Similarly desirable outcomes are obtained when carrying out the proposed SRSM on the whole City of Quebec (total area of 656 km2), yielding an area-based Quality of 62.37% and an object-based Quality of 63.21%.

Simulating realistic sonar data is crucial for tuning detection and classification algorithms according to environment and acquisition characteristics. Moreover, robustness of performances estimation and prediction applications can be greatly enhanced as soon as such a simulation tool provides both a modular underwater world representation (multiple sensors, environments and acquisition conditions) and a selection of several computational engines (ray theory, parabolic equation ... ). Therefore, we developed such a framework for simulators, allowing both scene design and computational engine choice. Within it, two engines (one for rays, one for tubes) has been successfully implemented and realistic simulations obtained, as shown in the presented simulated sonar images. A further step will be to also output the full received acoustic signal.

No abstract

Automatic extraction of buildings in urban scenes has become a subject of growing interest in the domain of photogrammetry and remote sensing, particularly with the emergence of LiDAR systems since mid-1990s. However, in reality, this task is still very challenging due to the complexity of building size and shape, as well as its surrounding environment. Active contour model, colloquially called snake model, which has been extensively used in many applications in computer vision and image processing, has also been applied to extract buildings from aerial/satellite imagery. Motivated by the limitations of existing snake models dedicated to the building extraction, this paper presents an unsupervised and automatic snake model to extract buildings using optical imagery and an unregistered airborne LiDAR dataset, without manual initial points or training data. The proposed method is shown to be capable of extracting buildings with varying color from complex environments, and yielding high overall accuracy.