A three-dimensional image processing method based on data layered two-dimensional normalization

Three-dimensional high-resolution imaging is a new technology for long-range detection of underwater small targets, This technology enables three-dimensional mapping of targets in water, sinking and buried environments. The three-dimensional imaging sonar adopts the ultra-wideband low-frequency emission array with large glancing angle and the two-dimensional acoustic array with wide directivity reception, which effectively acquires underwater target depth dimension, navigation dimension and azimuth dimension information. The pulse compression processing, synthetic aperture tomography processing, and anti-reverberation high-resolution array processing are used to realize three-dimensional high-resolution imaging of underwater targets. However, since the underwater echoes come from different azimuths, depths, and media environments, the significant difference causes the reverberation of the target image area to be severely unbalanced, which directly affects the background brightness and feature contrast of the target image area, and ultimately increases the artificial identify the risk of errors. In this paper, A three- dimensional image processing method based on data layered two-dimensional normalization is studied. Firstly, by tracking the three-dimensional seabed interface, the distribution law of seafloor reflection echo energy in different azimuths is statistically obtained. Then, the acoustic energy attenuation loss of different geological strata is compensated. After that, the dynamic range of the three-dimensional acoustic image and the target feature contrast are improved by the two-dimensional normalization processing of the layered water and the stratum. The method has been processed and verified by experimental data, so as to achieve stable and efficient realization of 3D acoustic image background equalization and target enhancement, which has strong robustness, certain innovation and good engineering use value.