6. Conclusions

In this chapter, the performance of a developed vision-based tactile detection algorithm, which was used to recognize the shape of tactile pavement for navigation purpose, was evaluated. The vision-based tactile detection algorithm was proposed and the experimental study on effectiveness of the detection algorithm by using five phases, which are load image, preprocessing, parameters extraction, metric calculation, and auditory output, was conducted. The proposed vision-based tactile paving detection system was also confirmed to be functioned in order to differentiate variety of shapes such as circle, bar, eclipse, square, triangle, and diamond. All the metric values could be applied as benchmark metric values for the next step of development of visually impaired navigation system.
