Currently, unmanned systems are seeing rapidly and extensive integration into several businesses, encompassing sectors such as healthcare and transportation. In the field of precision agriculture, a range of unmanned systems, including mobile robots, Unmanned Aerial Vehicles (UAVs), and Unmanned Ground Vehicles (UGVs), are employed. This paper presents the concept of a completely autonomous vehicle capable of detecting vegetation, navigating around obstacles, determining optimal routes, and executing independent movement. The utilization of a stereo camera is imperative for the execution of various processes, encompassing the important mapping process. All processes, including the mapping process, are conducted with a stereo camera model. This research centers around an autonomous ground vehicle designed to perform mapping and agricultural spraying tasks. Notably, the vehicle is developed to fulfill its primary function without relying on 3D LIDAR or GPS technology. Within the scope of this investigation, a model is trained using the original dataset, which comprises two distinct classes of yolov7: wild and cultivated plants. The training process yields a model that exhibits a precision rate of 99.6%. Furthermore, the identification of obstacles is accomplished through the utilization of depth information obtained from the camera system installed on the vehicle. By utilizing the IMU sensor included within the camera and incorporating depth information, a two-dimensional array is employed to replicate the surrounding world. Through the examination of the incoming data pertaining to the simulated environment, as well as the positions of vehicles and obstacles, it becomes feasible to determine the appropriate placement of vegetation and agricultural products, whether they have been subjected to spraying or not. Ultimately, the autonomous determination of the vehicle's probable route can be determined.