there is some anomaly than can be identified in well log absolutely such as the differences of gamma ray, resistivity quantitive, and density. In seismic interpretation, the contacts almost undetected because resolution of data (based of my knowledge, one period of seismic wave represent 30 meters of outcrop in subsurface). Seismic interpretation just allow us to detect the distribution of fluids and make geomodelling of reservoir in 3D mode. but, in explore the fluid contacts, we have to combine it with well-log data or core samples.

Simplifying the case in seismic interpretation, Reflection Coefficient (RC) between two different rocks depends on the difference between their Seismic Impedance that mostly depends on P-wave velocity in the rock. For example, the contact between a sand with water saturation and a sand with a certain gas saturation gives a high RC that produce an amplitude anomaly in the seismic trace, than it is possible to distinguish a flat horizon constituted by the aforesaid contact.

I attached an image of a flat spot (from http://geoseismic-seasia.blogspot.it/2008/11/baram-basin.html)

The terms oil-water and gas oil contact are misleading to an extent as they refer to transition zones rather than surfaces. The height of each transition zone is a function of relative permeability of the fluids, capillary pressure, and reservoir saturation. Oil - water transition zones can range from a meter to hundreds of meters and can vary in thickness from place to place in the reservoir.

Tim Schowalter published a good reference for understanding the subject in the 1979 AAPG Bulletin titled 'Mechanics of Secondary Hydrocarbon Migration and Entrapment', pages 723 - 760. A good secondary reference is By Schowalter and P. D. Hess, 1982, 'Interpretation of Subsurface Hydrocarbon Shows' also in the AAPG Bulletin at pages 1302-1327.

For well logs, Schlumberger has several free publications on interpreting hydrocarbon transition zones from well logs.

In seismic data, gas -water and sometimes oil water contacts are identified by bright spots as indicated above. These are just one class of attenuation anomalies or AVO's. CGG has a AVO tutorial website, but a search will find several good papers on Amplitude Anomalies.

As for other, core data and sample data are useful. However, in examining cores and samples, what we see at the surface at the surface are the conditions of the reservoir when it has been completely produced to surface pressures. This presents the possibility of confusing a residual reservoir with a current reservoir. Again I recommend reading the Schowalter articles to appreciate the necessity of multiple data sources.

After some thought, I find Mr. Bikshala's question is what I believe to be a fundamental question in oil & gas exploration. While I still recommend Schowalter's articles, I would also suggest that he look at a presentation that I made at the inaugural meeting of the Duliajan chapter of APG in 2010. I have posted a PDF file of the slides and authors comments on this website.

Look for Brown's write up on Seismic attributes.. These contacts come under the name DHI's which include Bright spots, flat spots, phase reversals. On well logs, you need to watch for an hourglass signature formed by both Neutron and Density in your reservoir interval to identify GOC and OWC.

For 3D seismic data, direct hydrocarbon indicators such as flat spots are great to find. You'll need high quality seismic to image those. If you have angle stacks, amplitude versus offset differences between near, mid, far, and ultra-far stacks can suggest both hydrocarbon type and contacts. Others have elaborated on those above. As Pier mentions, imaging a contact is all about understanding the relative difference in impedance between the two intervals. Bigger the impedance difference, the "brighter" the seismic trace will be.

For well logs, the resistivity traces can show changes at the OWC depending on the salinity of the water. For saline waters (which are conductive), the water baseline will sit to the left (lower ohmm reading) of the shale baseline and, obviously, well below that of the hydrocarbon bearing reservoir. Neutron and density tracks can also show contacts and suggest hydrocarbon type as others have mentioned above.

Direct hydrocarbon indicators (DHI)  also by using AVO and seismic attributes you can see bright spots , you can identify the OWC by well logs as Resistivity and sonic/Neutron logs together.