<p>This paper presents a lossless compression and recovery method for locally dense sparse photomask (LDSP) for ultraviolet lithography using irregular block-compressed row storage (IBCRS), which achieves an equilibrium reinforcement design for both storage amount and storage space. According to Moore’s Law, in laser three dimension printing (3DP), as well as in the semiconductor and chip manufacturing fields, the fabrication resolution is continuously refined for ultra-high-resolution manufacturing of complex curved surface components, resulting in an exponential increase in the generated metadata in terms of storage amount and storage space. As the first step, a photomask, which is a large-scale sparse matrix (LSM), is generated from variously sized irregular 3D manifolds using adaptive multilayer slicing strategies. Based on the typical lossless block compressed row storage (BCRS) method, irregular rules are proposed to tackle LSM, whose domain is locally dense via a neighborhood topology. The Hough transform of LSM was employed to determine the optimal transpose for the IBCRS. Using the IBCRS, the photomask of the LSM can be converted into lossless equivalent vectors. Large-scale matrix computations can be directly performed using compressed vectors. Original native photomasks can be recovered or reconstructed solely from compression vectors. The numerical example proves that in terms of the storage amount, which reflects the time complexity of the algorithm, the proposed IBCRS method performs better than the BCRS. Finally, a physical experiment was conducted to validate the IBCRS via the digital light processing (DLP) technique. The experimental results proved that the IBCRS had important applications in manufacturing variable-size devices with high precision, on the millimeter, micron, nano, and even atomic scales, by image-like lossless compression in the case of wireless encryption for space 3D printing, extremely high resolution manufacturing, etc.</p>