Unmanned surface vehicle is type of ship that can be controlled by a device. This paper proposes a custom object detection algorithm named FAP-YOLO v8x based on YOLO v8 with additional three modules to support the algorithm from issues like angular periodic problem, blurred image, false detection, no detection. The first module is De-pixelator module which reduces the noise level and upscales the image and returns the output to the next module. Second module is Angle detector module will analyze the state of image whether it is in necessary ascent or not. If not mathematical formula is used and the distance from center to edge of target is calculated and it is rotated to accepted ascent. Third module is the FP reducer module which supports the False positive rates, no detection, and false detection. The main module is YOLO v8x object detection algorithm will take detection USV with high accuracy and precision. The YOLO v8 algorithm has head module and backbone module to detect faster. The proposed algorithm gives accuracy of 96% and precision of 98.76%.