A "scientific" moral philosophy is still rooted in some fundamental non-objective rule, of course, but that's not necessarily a bad thing nor does it affect its plausibility.
One fundamental is that humans are separated from other animals[1] not just by self-awareness but also awareness of others — and the resulting effect of empathy. With this information, it can be argued that this ability bestows an obligation to care for others. A consistent moral philosophy can be built from that and other observations.
[1] This could apply to equal or lesser degree to some other species but we have no way of verifying it currently.
Here's my suggested reading list: Plato to establish a foundation, Descartes's First Meditation (edit: or the skeptics, but I don't really feel the need to subject non-philosophy students to that) to tear it down, Kant to give you a more rounded view (but be careful to note where Kant fails as well), Hume to bring reasonableness into being, Nietzsche to round out the pitfalls of nihilism, and some Neo-Kantianism, evolutionary psychology, and modern ethics to round it out.
How can you conceal the fact that you're trying to extract a conclusion from a system that doesn't have a premise? Well, there are indeed lots of methods of misdirection.
The last try at building a perpetual motion machine that I remember involved claiming to have a device for "amplifying power" but not being able to hook the device to itself "somehow". Made the system harder to disprove.