In this study, the antenna beamwidth filter (ABF) was used to eliminate the image noise and clutter originating from the back-projection algorithm (BPA) used in near field ground-based synthetic aperture radar (GB-SAR) imaging. Since BPA interpolates the range profile obtained in each synthetic aperture onto the region to be imaged, it spreads the scattering collected in the beamwidth of the antenna to the whole region and as a result, noise occurs in the focused image. Thanks to the simple but effective technique presented in the study, each range profile was spread only to the area where the antenna beam corresponds to, and a noise-free and very clean radar images have been obtained. The effectiveness of the ABF was first verified with the simulation data and then two different real experiments were carried out. Both simulation and real measurements were carried out with step frequency continuous wave radar form. In the first experiment, because the synthetic aperture was relatively short and there was no dominant scatterer other than a metal reflector in the scene, very high image distortions did not occur between the traditional BPA and the filtered BPA results. Dispite these results, there were visually noticeable distortions that occurred in the focusing of the target and noisy signs appeared in the imaged scene. In the second experiment, quite notable differences were observed between filtered and traditional BPA results, as the scan was performed over a relatively long distance and there were too many scatterers in the target area. In the image reconsctructed using filtered BPA, while all targets were focused with high resolution and high contrast, in the image obtained by traditional BPA, it was observed that some small targets disappear under the noise signs. As a result of all these findings, it has been revealed that BPA with ABF has quite high performance and image quality.