The use of efficient signal processing tools (SPTs) to extract proper indices for the fault detection in induction motors (IMs) is the essential part of any fault recognition procedure. The 2nd part of this two-part paper is, in turn, divided into two parts. Part two covers the signal processing techniques which can be applied to non-stationary conditions. In this paper, all utilized SPTs for non-stationary conditions have been employed in details for fault detection in IMs. Then, their competency and their drawbacks to extract indices in the transient state modes are investigated from different aspects. The considerable experimental results are given to certify the present discussion. Different kinds of faults including eccentricity, broken bar, and bearing faults as major internal faults in IMs are investigated.