An architecture for visual motion perception of a surveillance-based autonomous robot

Abstract

This research presents an innovative mobile robotic system designed for active surveillance operations. This mobile robot moves along a rail and is equipped with a monocular camera. Thus, it enhances the surveillance capability when compared to conventional systems (mainly composed by multiple static cameras). In addition, the paper proposes a technique for multi-object tracking called MTMP (Multi-Tracking of Motion Profiles). The MTMP resorts to a formulation based on the Kalman filter and tracks several moving objects using motion profiles. A motion profile is characterized by the dominant flow vector and is computed using the optical flow signature with removal of outliers. A similarity measure based on the Mahalanobis distance is used by the MTMP for associating the moving objects over frames. The experiments conducted in realistic environments have proved that the static perception mode of the proposed robot is able to detect and track multiple moving objects in a short period of time and without using specialized computers. In addition, the MTMP exhibits a good computational performance since it takes less than 5 milliseconds to compute. Therefore, results show that the estimation of motion profiles is suitable for analyzing motion on image sequences.

A mosaicking technique for object identification in underwater environments

“…Nowadays, intelligent and autonomous robotic vehicles are developed with visual systems (Pinto et al , 2014a, 2014b, 2014c) to obtain visual evidence of the sea-floor, and as a consequence, researchers have proposed some techniques to map large areas. Pizarro and Singh studied the impact of the reduced overlap between consecutive images (causing a radial distortion), unstructured motion and the poor lighting conditions.…”

“…Therefore, it becomes crucial to develop perception methods that make underwater seafloor mapping possible whenever the vehicle moves closer to the bottom of the sea. These methods will allow a more efficient use of such systems in real environments and applications (Pinto et al , 2014a, 2014b, 2014c). Maps generated by mosaicking methods can be used to object detection and identification, in particular, to discover sinked structures and to interpret the biological and physical characteristics of marine elements (Foresti and Gentili, 2002; Srividya and Shobha, 2014).…”

“…In the pre-processing phase (a blue block on Figure 1) intends to remove spatial-temporal noise from the input image. This technique is named robust bilateral and temporal (Pinto et al , 2014a, 2014b, 2014c) and uses the spatial and temporal evolution of the sequences to reduce the noise component and to make the images more appealing. It is quite common for the presence of suspensoids in the water column, which can be considered as noise component.…”

Altitude control of an underwater vehicle based on computer vision

“…For the implementation of this feature the software presented in [41] was used. It allows the configuration of the parameters of the camera, for example, frames per second, resolution, and exposure time.…”

“…After filling the array mentioned, this array is sent via UDP communication. This functionality was implemented by modifying and adding necessary methods to a class present in the software used in [41] already mentioned, class named Protocol. This class originally allowed the features of sending and receiving UDP packages.…”