We present an approach for improving spatial frequency sampling in active incoherent millimeter-wave (AIM) imaging systems using frequency diversity. AIM imaging relies on active transmission of spatio-temporally incoherent signals to illuminate a scene, from which interferometric Fourier-domain imaging can be implemented using a sparse receiving antenna array. One of the benefits of Fourier domain imaging is the sparsity of the receiving array, which can form images with equivalent resolution to traditional filled beamsteering arrays, but with a small fraction of the elements. The hardware reduction afforded by the sparse array often leads to an undersampled Fourier space, where even though image formation is possible, the image reconstruction may be degraded when viewing complex objects. To address this challenge without requiring additional receiver channels, we explore the use of frequency diversity in the illuminating and receiving systems. Fourier domain spatial frequency samples are determined by the electrical spacing and rotation of the receiving elements, thus by changing the frequency the sampled spatial frequencies also change. We implement an additive technique where the spatial frequency samples are summed prior to Fourier transform image formation. Importantly, because the system is active, a consistent signal-to-noise ratio is maintained across all frequencies, which may not be possible in traditional passive Fourier-domain imagers.