Detection of rockfill gradation based on video image recognition

The gradation properties of earth-rock dam-filling materials primarily influence their mechanical properties after compaction. Herein, combined with the characteristics of rockfill gradation detection (RGD) requirements in practical engineering, an intelligent detection method for rockfill gradation based on video image (VI) recognition technology is proposed and applied to engineering applications. By designing a device for VI recording, multi-group RGD tests are conducted. Further, a method for rockfill identification and particle size gradation transformation based on VI is proposed. The accuracy and error of the RGD in traditional screening and VI recognition method are compared and analyzed. The results of the laboratory tests and engineering applications in a construction site reveal that the proposed RGD method is highly efficient, precise, and robust.

Automated summary

In this study, an intelligent detection method for rockfill gradation based on video image recognition technology is proposed and applied to engineering applications. The proposed method shows high efficiency, accuracy, and robustness in laboratory tests and engineering applications, and outperforms the traditional screening and video image recognition methods.