Wireless health monitoring devices are proliferating in the current era. Ambulatory electrocardiogram (ECG) devices are one of the tools that help in long-term ECG measurement. Ambulatory electrocardiogram is a recent technology for analyzing heart activity from extensive ECG measurements. Generally, the ECG signals get tempered through different noises during measurement. Among the noises, motion artifacts and baseline wander are the predominant noises present in ambulatory ECG signals. It is a challenging task to remove these noises from ambulatory ECG recordings. This chapter presents a robust approach for motion artifact and baseline wander suppression using empirical wavelet transform (EWT) and non-local means (NLM) estimation. This denoising technique is tested using the Xilinx PYNQ-Z2 board. The efficacy is assessed using different parameters, i.e., increment in signal-to-noise ratio, correlation coefficients, and root mean square error. The efficacy of the proposed method is validated using MIT-BIH arrhythmia and MIT-BIH NSTDB database. The denoising technique discussed in this chapter performs better in suppressing motion artifacts and baseline wander from the noisy ambulatory ECG measurements. The verified hardware compatibility for the proposed algorithm with the PYNQ-Z2 board substantiates that the proposed technique can be implemented over an ambulatory ECG recording device for wireless ECG monitoring.