One of the basic functions of individuals with special needs living integrated with society is the appropriate urban infrastructure. Especially for visually impaired individuals, walking surface signs, that is, the condition of visually impaired roads, are extremely important. It is aimed to make travels safe for individuals by providing convenience in terms of guidance. Additionally, there is a need for technological developments that will further increase the integration of these individuals into society. Developing systems that can provide guidance and warning about walking surface signs with the help of head-mounted and glasses-integrated cameras or sensors will provide an advantage to these individuals. Considering the application of this technology, the sample segmentation process of visually impaired roads was carried out with the YOLOv8 object detection algorithm within the scope of the study. Training processes were completed with data obtained from different regions and high confidence values were obtained.