paper presents a new technique for calibrating heliostat errors which affect the heliostat efficiency and the production of thermal energy from the reflected beam from mirror to the target. This technique is based on analyzing the image for identifying different deterministic error types in single facet small heliostat, such as pedestal tilt and deformation due to wind loading. The technique depends on the image constructed from reflected beam. For each image, seven features are extracted and stored. These features are based on the moment invariant technique, which is insensitive to translation, rotation, and scale changes. The obtained results verify accurate and fast operation for heliostat calibration. The computed false negative rate is 0.06 and the recognition rate is 94.0476 %.