Laser speckle interferometry is used to detect micro-structure and its dynamic behavior of a sample surface by the statistical analysis of its laser speckle images. In this paper, the methodology of laser speckling and statistical analysis are studied with the aims to develop a non-contact, non-destructive surface sensing technique which potentially devices a compact, cost-effective tool for measuring the biological status of a plant by scanning its leaves. First, an auto-correlation based analysis method is proposed for the discrimination of various surface roughness levels using their laser speckle statistics. Second, techniques for dynamic speckle pattern analysis for the detection of the evolution of a time-varying sample surface are discussed. The effectiveness of proposed measurement methods are demonstrated via the experiment on a detached leaf.