• Uday Bheem Sai Chagantipati
• Susmitha Gurram

Autonomous vehicles produce and maintain a chart of the surrounding area grounded on the various sensors found in a different region of the vehicle. Radar sensors monitor the position of near vehicles. Videotape cameras detect traffic lights, read traffic signals, track other vehicles, and monitor pedestrians. Light detection and ranging sensor (Lidar) maps into the surrounding area of a vehicle to determine distances, track edges, and identify path signals. In autonomous vehicles, technology image segmentation proved to be a major problem in visual perception. Autonomous vehicles are being used in various applications, including manufacturing, hazardous materials handling, and surveillance. Image segmentation is applied in the visual perceptual functions of observing agents in the environment, identifying road boundaries, and tracking road signals. The purpose of image segmentation is to divide an image into several parts, each representing a different entity. The main aim of this paper is to make the input images into segments using the image segmentation process and Convolution Neural Network method for efficient results of visual perception. The proposed image segmentation method plans to standardize and promote the development of state-of-the-art methods for visual inspection system understanding by using a data-set sample and validation using Python language. Based on the experimental results, an NVIDIA GTX 1050 GPU achieves 73% mean IOU and 90 FPS inference speed.

Image Segmentation, Deep Learning, Autonomous Vehicles, Convolution Neural Network, k-means clustering.

"IMAGE SEGMENTATION BASED ON MACHINE LEARNING FOR AUTONOMOUS VEHICLES ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 12, page no.a136-a141, December-2021, Available :http://www.jetir.org/papers/JETIR2112017.pdf

"IMAGE SEGMENTATION BASED ON MACHINE LEARNING FOR AUTONOMOUS VEHICLES ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 12, page no. ppa136-a141, December-2021, Available at : http://www.jetir.org/papers/JETIR2112017.pdf