In this Letter, we propose a novel compact optical system for automated cell identification. Our system employs pseudo-random encoding of the light modulated by the cells under inspection to capture the unique opto-biological signature of the micro-organisms by an image sensor and without using a microscope objective lens to magnify the object beam. The proposed instrument can be fabricated using a compact light source, a thin diffuser, and an image sensor connected to computational hardware; thus, it can be compact and cost effective. Experiments are presented using the proposed system to identify and classify various micro-objects and demonstrate proof of concept. The captured opto-biological signature pattern can be attributed to the micro-object's morphology, size, sub-cellular complex structure, index of refraction, internal material composition, etc. Using the captured signature of the micro-object, we extract statistical features such as mean, variance, skewness, kurtosis, entropy, and correlation coefficients for cell identification using the random forest classifier. For comparison, similar identification experiments were repeated with a digital shearing interferometer. To the best of our knowledge, this is the first report on automated cell identification using the proposed approach.