HSV Semantic Segmentation on Partially Facility and Phanerophyte SunShine-Shadowing Road

EnRDeA U-Net Deep Learning of Semantic Segmentation on Intricate Noise Roads

“…This may be compared with the scavenger behavior of an herbivorous animal when it seeks out and consumes dead plants for feeding [4]. However, for an SDSB to fully achieve scavenger feeding behavior, many challenges must still be met in terms of machine vision for target detection [5][6][7] and road segmentation outdoors [8,9]. In a modern city, road conditions present a much more complex environment; this includes the presence of many interruptive noises, and the additional factors of changing weather and climate, as well as different kinds of road materials [8].…”

“…The road surface surroundings were considered with respect to intricate noises including differing weather conditions (sunny, rainy, or cloudy), sunshine spots and shadowing caused by large trees or physical facilities, traffic signs, obstacles, and types of road materials (cement or asphalt). The previous work investigated the morphological filtering on the road segmentation [9][10][11][12], in which [9] an optimizing HSV encoding framework is proposed by embedding the morphology operation to explore the road segmentaton. However, during the morphological filtering on segmentation, the different times of corrosion and expansion operatons were iteratively tuned subjectively to reach the optimized results, which implied that the different road optimized segmentation would rely on specific times of corrosion and expansion operatons through user inspection; this fact further indicated that to dynamically tune the parameters on corrosion and expansion operations for road segmentation is infeasible, in terms of the real-time self-driving applications.…”