Advanced Driver Assistance Systems (ADAS) Based on Machine Learning Techniques for the Detection and Transcription of Variable Message Signs on Roads

Among the reasons for traffic accidents, distractions are the most common. Although there are many traffic signs on the road that contribute to safety, variable message signs (VMSs) require special attention, which is transformed into distraction. ADAS (advanced driver assistance system) devices are advanced systems that perceive the environment and provide assistance to the driver for his comfort or safety. This project aims to develop a prototype of a VMS (variable message sign) reading system using machine learning techniques, which are still not used, especially in this aspect. The assistant consists of two parts: a first one that recognizes the signal on the street and another one that extracts its text and transforms it into speech. For the first one, a set of images were labeled in PASCAL VOC format by manual annotations, scraping and data augmentation. With this dataset, the VMS recognition model was trained, a RetinaNet based off of ResNet50 pretrained on the dataset COCO. Firstly, in the reading process, the images were preprocessed and binarized to achieve the best possible quality. Finally, the extraction was done by the Tesseract OCR model in its 4.0 version, and the speech was done by the cloud service of IBM Watson Text to Speech.

Visual Perception and Understanding of Variable Message Signs: The Influence of the Drivers’ Age and Message Layout

Variable message signs (VMS) are used to display messages providing up-to-date traffic-relevant information so that drivers can safely adapt their behavior in real time. The information reported in a VMS should be brief but comprehensive to minimize perception time. The latter can be influenced by the way the message is displayed. This study investigates how the different ways of displaying the same message can influence reading time and the information perception process at different driving speeds. Specifically, the following message characteristics are investigated: (i) use of uppercase and lowercase letters; (ii) use of familiar pictograms; and (iii) use of less familiar pictograms. Furthermore, as perception time typically changes with ageing, drivers belonging to three different age classes are tested. The experimentation was performed by simulating a vehicle passing along a straight road upon which a VMS displaying different messages was placed. Experimentation results are analyzed using the Kruskal–Wallis test, Friedman rank-sum test and Welch one-way ANOVA, showing that: (i) the use of uppercase or lowercase does not seem to significantly affect reading times; (ii) the use of pictograms that are not very familiar to habitual road-users can be counterproductive for the perception process; (iii) elderly drivers always have greater difficulty in perceiving the message than young or middle-aged drivers. The findings of this study can be of help for traffic authorities to design the most suitable structure for a VMS so that its information can be unequivocally and immediately conveyed to drivers.