Amaury Flancquart

Quantification of GNSS signals accuracy: An image segmentation method for estimating the percentage of sky

This paper is focused on the characterisation of the GNSS reception signals environment by estimating the percentage of visible sky. The estimation is based on a new segmentation technique that uses color and texture gradients with an adaptive and non-parametric combination strategy. The structural gradient, resulting from the combination, is processed with the watershed algorithm to get image segmentation. The classification process used to extract the sky region is performed using the k-means algorithm. Experimental segmentation and classification results, using real data and an evaluation methodology, are presented to demonstrate the effectiveness and the reliability of the proposed approach.

Adaptive model for object detection in noisy and fast-varying environment

This paper presents a specific algorithm for foreground object extraction in complex scenes where the background varies unpredictably over time. The background and foreground models are first constructed by using an adaptive mixture of Gaussians in a joint spatio-color feature space. A dynamic decision framework, which is able to take advantages of the spatial coherency of object, is then introduced for classifying background/foreground pixels. The proposed method was tested on a dataset coming from a real surveillance system including different sensors installed on board a moving train. The experimental results show that the proposed algorithm is robust in the real complex scenarios.