PROJECT TITLE :

A novel obstacle detection method based on deformable grid for the visually impaired

ABSTRACT:

To assist visually impaired individuals, a selection of obstacle detection (OD) methods with the monocular vision are developed. Most conventional OD methods detect the obstacle by using the variation of the motion vector or the article size between 2 consecutive image frames. However, the OD performance using the short-term variation is significantly plagued by the tracking error, which ends up in inaccurate OD results. To deal with this drawback, this paper presents a completely unique OD technique based on a brand new structure, known as Deformable Grid (DG). The DG is initially an everyday gridshaped, however it can be deformed gradually relying on the motion of the article within the scene. The proposed method detects the object at risk of collision based on the degree of the deformation of the DG. Experimental results demonstrate that the proposed OD method outperforms the traditional ones in terms of processing time plus accuracy.