Inferring 3D image from 2D image is an advance topic in computer vision. This article considers a 2D image depth estimation of an object and reconstructs it into a 3D object image. The 2D image is defined by slices, where each slice contains a set of points that are located along the object contour and within the object body. The depths of these slices are estimated using the neural network technique (N.N.), where five factors (slice length, angle of the incident light and illumination of some of points that located along the 2D object, namely control points) are used as inputs to the network. The estimated depths of the slices are mapped into a 3D surface using the interpolation technique of the Bezier Spline surface. Our model was tested and evaluated using different objects with different and complex shapes. The results showed an effective performance of the proposed approach.