Noise always influences on the result of signal-detection, therefore it is a hot topic to detect signal under the low signal-to-noise ratio (SNR) in current detection field. However, it is an intractable problem of signal-detection to extract weak signal under a high noisy environment. Adaptation filter provides a simple and useful way to detect signal. Through measuring and studying, it is possible to carry out the best combination of weak signal under the condition of weak signal. In this paper, we propose a novel mechanism to detect the ECG signal based on adaptation wavelet transform (WT). We adopt the sub-strap coded theory of WT and restore post-combination weak signal through adaptation matching of multi sub-straps' weights. The simulation results show that the proposed method can further improve the ability to detect signal, which is a very effective scheme to detect the property of weak signal and improve SNR.