This paper exploits laser speckle photometry (LSP), a full-field non-contact optical-based image analysis technique, for effectively and rapidly imaging hidden damage in structures, rather than with the complex setups in digital speckle pattern interferometry (DSPI) or shearography (SG)-based interferometry. This technique will demonstrate a promising potential for large-area inspection of composite structures in near real time to unearth barely visible impact damage (BVID) which would typically go unnoticed during routine inspections. Three image (processing) algorithms for localizing and then imaging the BVID area were explored: conventional mean squared error (MSE), normalized cross-correlation (NCC), and an index-centered algorithm known as structural similarity index measure (SSIM). When implementing these algorithms in LSP, a pre-processing step of selecting a window size (subregion size) for locally correlating the images to localize the damage and estimate its size was performed as an advancement on the previous pixel-by-pixel correlations made with LSP. A trade-off strategy between two perceptual-based metrics, image fidelity and image sharpness/blurriness, was implemented to evaluate an appropriate range for window sizes followed by the image algorithms for creating high contrast imaging of damage regions. From the correlation map, the strategy was carried out to mitigate image noise caused by the camera (image) sensors and speckle patterns dictated by the overall root-mean-square deviation (RMSD) while maintaining a high level of sharpness characterized by the magnitude of the third-level discrete wavelet transform. The proposed image algorithms in conjunction with the appropriately selected window size served as imaging conditions in the context of laser speckles for the first time and were tested on BVID in an impacted honeycomb composite panel under thermal excitation. A low coherence (high-power) laser for fast screening of a large area, if required, followed by a high coherence (low-power laser) for detailed imaging were used to demonstrate the efficacy of LSP. The damage image region agreed well with a baseline image from the well-established point-by-point CT-scan with all the three image algorithms. Overall, NCC and SSIM performed slightly better than MSE, with SSIM generally being the better of the two. Nevertheless, MSE has its merits with ease of interpretation and implementation. LSP with the proposed imaging conditions shows enormous potential as a real-time non-destructive inspection (NDI) technique not only in the aerospace industry but also in industries such as additive manufacturing where on-line in-situ monitoring is desired for prevalent defects. The real-time inspection using LSP will further allow immediate feedback for process controls.