To achieve automated detection of liquor-making raw materials during the storage process, enhancing the efficiency and objectivity of storage, an automated online inspection system was developed. The system primarily consists of modules such as automatic sampling devices, automatic feeding and distribution systems, machine vision detection, and near-infrared detection. It automates processes like sampling, detection, material return, and detection information transmission. This system enables the automated inspection of sorghum and wheat for unsound kernels, moisture, and starch indicators during the storage acceptance process. Validation results indicate that the system's average omission rates for sorghum and wheat unsound kernels are 0.34% and 0.12%, respectively, while the average false detection rates are 0.37% and 0.16%. The maximum SEP for moisture and starch indicators in sorghum and wheat is 0.96%, with a maximum Sr of 1.73%, demonstrating that the accuracy and stability of the system's detection indicators meet practical application requirements.