Abstract: The dolomite reservoirs of Majiagou Formation of the Ordovician series in the central and eastern part of the Ordos Basin are characterized by poor physical properties and strong heterogeneity. The signal-to-noise ratio (SNR) of nuclear magnetic resonance (NMR) logging data and the accuracy in porosity calculation are low, leading to difficulties in the evaluation of reservoir effectiveness. A singular value decomposition (SVD) filtering method based on NMR echo optimization was proposed, as well as a porosity correction approach incorporating gas-bearing light hydrocarbon correction, which reduced the average relative error in porosity calculation by approximately 20%. Moreover, reservoir classification criteria for the study area were established. Firstly, a SVD filtering method based on NMR echo quantity optimization was established to preprocess the echo data with low SNRs. Then, NMR logging and density logging were employed for light hydrocarbon correction, significantly improving the accuracy of porosity calculation in the study area. Finally, reservoir classification criteria were established using calibrated NMR porosity and resistivity parameters for different fluid types. This research provides reliable technical support for the exploration and development of low-porosity reservoirs in the Ordos Basin, effectively promoting further hydrocarbon exploration and production in the study area.