The work aims to accurately locate the wrinkle features of the existing corrugated board packaging, and ensure that the wrinkle features of corrugated board packaging containers can be timely and effectively inspected in production. According to image information collection, recognition algorithm, RGB vision, location calibration, convergence result comparison, etc., the machine vision corrugated board packaging inspection system determined the key algorithm points of corrugated board packaging wrinkle feature calibration. The RGB visual improvement algorithm was analyzed based on the traditional frequency-domain transformation. The wrinkle feature recognition rate of corrugated cardboard packaging was effectively improved. The testing points of this kind of packaging were clarified. With the actual corrugated board as an example, the feature extraction method of the wrinkle features of the packaging was demonstrated, and the visual features of the wrinkles were strengthened. Finally, the convergence results were compared. The linear rate of passing the threshold of 90~120 was 25%~80%. The results showed that this vision method can realize the corrugated board packaging calibration of machine vision, and the actual regression base rate of each feature was 86%, verifying the accuracy of RGB vision algorithm for wrinkle feature calibration. Based on the RGB vision and optimization measures, the machine vision calibration method of corrugated board packaging is determined, which meets the effective positioning requirements of corrugated board packaging production.