For large astronomical telescope system based on adaptive optics(AO), the correcting ability of the wavefront processor in real time is the critical factor to the target observation. However, when we are observing the faint-target, it will lead to the uneven energy distribution of the spot in the Hartmann sensor, even no spots in some of the sub-apertures. However, whether to control the adaptive optical system to close loop is mainly dependent upon the distribution of spot in the microlens of Hartmann sensor, so it is necessary to detect the existence of the sub-spot in the Hartmann sub-aperture. This paper presents an improved approach based on template matching and threshold processing to detect the sub-spot, it could judge whether there is spot in sub-aperture self-adaptively, then, we provide an intelligence foundation for adaptive optical system. In this paper, not only do we give the detailed implementation of the detection algorithm, but also test it with the reality observed images. The experimental results have shown that the proposed approach in the paper could accurately detect whether there is spot in Hartmann sub-aperture.