Abstract: Continuous monitoring of the condition of a rotating machine is an important and required task for engineers and researchers in industry. For any rotating machinery bearing is the core element. For that reason health monitoring of bearing in a rotating machinery is very important. Vibration is one of the most widely used signature used for the health monitoring of the bearings. In this research, the experiment is executed in two stages. As the vibration signal acquired from the bearing set-up is in general noisy in nature, so in the first phase of the experiment, the noise present in the vibration signal is removed to improve the SNR. This noise filtering is done using the ANC (Adaptive Noise Cancellation) technique. Initially, three ANC techniques are employed on the vibration signal acquired from the experimental set-up. The performance of the ANC techniques are compared. From the comparison EMD is found better. So, EMD algorithm is used for the implementation of the adaptive noise cancellation in the preprocessing of the vibration signal and then the filtered signal is used in the next phase of the experiment for further analysis to detect the bearing defect. As the time domain (static analysis) or frequency domain analysis alone may not provide the precise information about the defect, so in the second phase of the experiment the static analysis and the frequency domain analysis along with the time-frequency analysis is done on the filtered vibration signal to identify the defect in the bearing.