In NiFe/Pt bilayer, when spin current originating from the magnetization procession of NiFe is inject into the adjacent Pt layer under ferromagnetic resonance (FMR), the direct current (DC) voltage VISHE generated by inverse spin Hall effect (ISHE) will be added to the voltage VSRE generated by spin rectification effect (SRE), therefore the measured voltage in experiment is the sum of VISHE and VSRE. It is crucial to separate these contributions, which has been often overlooked before, in order to make a reasonable comparison of the ISHE among different materials. The voltages having symmetric (Lorentz type) and anti-symmetric (dispersive type) components both vary with the static magnetic field strength. However, they have different static magnetic field angle dependences according to our theoretical analysis. In order to distinguish the contribution of ISHE from that of SRE, in this paper, we employ a method, in which the voltage across the sample is measured when the static magnetic field is applied to different directions, to analyze the voltage by varying magnetic field angle in a range from 0° to 360° in steps of 10°, thereby separating the VISHE. The separation is carried out by fitting the angle dependent symmetric and anti-symmetric curves to different theoretical formulas of ISHE and SRE. The voltages of the two different contributions together with the phase angle of the microwave are obtained. At the same time, the FMR line width and the resonant field can be read out. The results show that the ferromagnetic resonance line width in NiFe(20 nm)/Pt(10 nm) sample is larger than that in NiFe(20 nm) sample due to the injection of spin current from NiFe to Pt in the bi-layer sample. We notice that in the curves of voltage vs. static magnetic field, the Lorentz symmetry components of the voltage from the bi-layer sample weight more than those from the single-layer sample. This is explained as a result of the existence of the ISHE in the bi-layer sample, where the spins are pumped from the magnetic layer to the adjacent nonmagnetic layer. The spin pumping effect does not show up in the single-layer sample. There are a large portion of symmetric components in the double layer sample, which is attributed to the ISHE. Although the voltage caused by the SRE is smaller than that by the ISHE, the SRE voltage cannot be ignored. Our work is crucial to understanding the spin-related effects in ferromagnetic/nonmagnetic metal material and provides an improved analysis method to study the spin pumping and the ISHE.