In this paper, a simple analysis method for flow simulations around a motor cycle is discussed based on CFD. In the practical CFD, most of the efforts are wasted for grid generations. The typical two types of grid are body-fitted grid and Cartesian grid. Body-fitted grid was often used for the flow calculation around the vehicle until today. However, grid generation around the complex object involves an immense amount of time and effort. Now, Cartesian grid is in focus. This approach is one solution to save time for grid generation. In the recent years, accuracy of Cartesian grid approach has been raised by increasing the number of grid points with the progress of computer. When computing of flow around the object, CAD data of the object is necessary. However, CAD data has not always had to be prepared. In the present approach, flow computation is done based on a digital image of a motor cycle with a person, instead of the CAD data which is generally required to represent the object shape. When analyzing aerodynamic characteristics of the motor cycle, human body data are necessary as well as that of the motor cycle. Unlike the motor cycle, human CAD data are not easily obtained. Besides the human takes different postures and they definitely affect to aerodynamic characteristics. In the present paper, we propose a new approach to incorporate the human body and the motor cycle shapes into Cartesian grid without any CAD data, but with digital images. To set the shape of a motor cycle in Cartesian grid, the object model is needed. A human body pose flexibly changes when a biker really rides a motor cycle. A different rider has a different body figure, as well. When human body pose and human body figure change, CAD data should be rewritten. In this research, digital photo is used as a substitute for CAD data. In order to insert this image into Cartesian grid, the object is clipped from the base image by image processing and is colored by parts, such as a frame, tires and so on. The resolution is reduced to fit the grid size. For the model of three dimensions, the object images are obtained from multidirections and the object is clipped and colored as the same as two dimensions. Pseudo-compressibility method is used for the incompressible flow simulation. Higher-order upwind difference is used for the convection terms and second-order central difference is used for the viscous terms. The temporal difference is first-order and Lower Upper-Symmetric Gauss Seidel (LU-SGS) employed for the convergence calculation of pseudo time. In this image-based method, flow around a motor cycle with a person and aerodynamic characteristics of the motor cycle were successfully obtained.