Conventional energy detection (ED) spectrum sensing method requires the knowledge of noise power and consequently suffers from noise uncertainty. In this letter we propose a novel blind spatial spectrum sensing method to conquer this problem. We utilize an Electronically Steerable Parasitic Antenna Receptor (ESPAR) which can divide the space into several sectors and switch receive beampattern to each sector in a time-division fashion. The difference of received signal energies among sectors heavily depends on the presence/absence of the primary user (PU) signal. Motivated by this fact, we introduce a blind energy-based spatial spectrum sensing algorithm using the ratio of maximum-minimum energy (MMEN). The MMEN spectrum sensing method can overcome the noise uncertainty problem and achieve higher accuracy performance with lower computational complexity. Theoretical performance analysis of the proposed MMEN method is also provided. Simulation results are presented to verify the efficiency of the proposed algorithm.