The aim is to detect lane and object with higher accuracy rate, even in different climatic conditions. Using Ross kippenbrock principle for finding lane (Advance lane detection) and YOLO for object detection which has a higher speed and accuracy rate. Self-driving cars might not be in our everyday lives yet, but they are coming. Analyzing images and figuring out where the lane lines are on a given roadway is one of the core competencies of any respectable self-driving car. After getting the output for the lane detection, these frames are sent to the CNN model using YOLO (you only look once) which will detect the object which are present in the frame. These two methods gives a higher accurate and spped result, which is must for self-driving car.