A novel active surface Laplacian electroencephalogram (LEEG) sensor for the real-time mu rhythms detection has been developed In our study. Analog LEEG signals with high signal to noise ratio obtained directly by using the active sensor can reduce the duration and quantum error of digital signal processing and computation for the quick and precise control of brain-computer interface (BCI) systems. The portable active surface LEEG sensor comprises five gold electrodes integrated to a cross-shaped structure and a battery-powered, low-noise amplifier with the following specification: gain of 10,000, band-pass filtering from 2.5 Hz to 55 Hz, input impedance of 10 GΩ, common mode rejection ratio (CMRR) of 110 dB. The clinical experiments showed that the amplitude suppression of mu waves detected directly by active surface LEEG sensors was obvious as the normal subject was asked to imagine grasping something with his right hand. Furthermore, the distribution of mu waves captured real time in the un-shielded room by active surface LEEG sensor was similar to that acquired in the shielded room of hospital through EEG data collection by an EEG instrument, and then offline analyses. Real-time mu rhythms obtained by the active surface LEEG sensor will be utilized to control a device or system via the BCI in real time.