This paper presents a novel ECG signal measuring approach using compressive sensing method. The signal representing sparsity in any orthogonal basis can be well recovered using minimize L1 norm optimization, while satisfying the RIP condition for the measurement matrix and orthogonal basis . First, based on this theorem, an analysis for evaluating the sparsity of ECG signal in orthogonal basis domain is proposed. A set of ECG samples from MIT-BIH medical database are adopted into Fast Fourier Transformation (FFT) process. The results indicate these signals can be well represented in sparsity using conventional orthogonal transforms. Second, the lightweight recovery algorithm is proposed based on orthogonal matching pursuit using iterative function and least squared approximation. The simulation results show that the special features in ECG including QRS complex, R-R interval, PQ and ST duration, can be well recovered with negligible norm error. It also indicates that using this approach can significantly save the total power of ECG acquisition.