The star CoRoT102781750 reveals a puzzle, showing a very complex and altering variation in different `CoRoT colours. We established without doubt that more than a single star was situated within the CoRoT mask. Using a search for periodicity as a too
l, our aim is to disentangle the composite light curve and identify the type of sources behind the variability. Both flux and magnitude light curves were used. Conversion was applied after a jump- and trend-filtering algorithm. We applied different types of period-finding techniques including MuFrAn and Period04. The amplitude and phase peculiarities obtained from the independent analysis of CoRoT r, g, and b colours and ground-based follow-up photometric observations ruled out the possibility of either a background monoperiodic or a Blazhko type RR Lyrae star being in the mask. The main target, an active star, shows at least two spotted areas that reveal a $P_rot = 8.8$ hours $(f_0 = 2.735$ c d$^{-1})$ mean rotation period. The evolution of the active regions helped to derive a period change of $dP/dt = 1.6cdot 10^{-6}$ (18 s over the run) and a differential rotation of $alpha = DeltaOmega/Omega = 0.0074$. The $0fm 015$ linear decrease and a local $0fm 005$ increase in the dominant periods amplitude are interpreted as a decay of the old spotted region and an appearance of a new one, respectively. A star that is detected only in the CoRoT b domain shows a $f_1 = 7.172$ c d$^{-1}$ pulsation connected to a $14fd 83$ periodicity via an equidistant triplet structure. The best explanation for our observation is a $beta$ Cep star with a corotating dust disk.
