Research On Underwater Lidar Detection Based On Adaptive Backscattering Filtering Technique

Lidar technology has been widely used in underwater target detection, recognition and ocean surveying and mapping. However, the light scattering and absorption of water seriously influence the detection performance of the underwater Lidar system. At present, the underwater Lidar range finder used to be mounted on airborne or shipborne platforms, both of them have disadvantage in mobility, volume size, operating cost, measurement area and time. Therefore, the automatic, intellectualized and small-sized underwater Lidar detection system, which can be mounted on unmanned underwater vehicle(UUV), becomes the urgent need. This paper presents a backscatter light adaptive filtering method, which combines the requirements of unmanned underwater vehicle equipment and the characteristics of underwater target Lidar echo and underwater detection environment. The performance under different absorption coefficient and scattering coefficient and the corresponding signal processing results are investigated by Monte Carlo simulating calculation. Finally, according to the theoretical research result, the underwater target detection Lidar prototype has been designed and completed. The detection performances and underwater task capability of the system prototype have been verified by the laboratorial and the natural water environment experiments.On the basis of the time of flight ranging principle, the influencing factors of the underwater target distance measurement range and precision are analyzed in detail. The impacts of the water inherent optical properties on underwater optical transmission are introduced. The underwater ranging lidar equation and water backscattering calculation model are derived by the physical definition of the water apparent optical properties in radiometry. According to the coverage requirement of the underwater detection beam, a focusing method based on phase modulation of azimuthally polarized light is proposed theoretically. Our proposed method achieves a small focusing spot, which provides the theoretical basis for the generation of large probe beam.According to the requirements of inhibiting water backscattering, the adaptive filtering technique based on the variable forgetting factor recursion least square(VFF-RLS) method is proposed. The innovative introducing of the forgetting factor function,which is related to the characteristic of underwater Lidar echo signal, strengthens the inhibition effect of water backscattering and the sensitivity to the weak target echo signal. Then, the influence of filter parameters on the signal processing results is discussed. We also provide the evaluation criterion of inhibition effect in water backscattering signal. Through comparison and analysis of timing error in different timing discrimination techniques, constant fraction discriminator(CFD) is more suitable for the underwater flight time measurement. And the selection of the optimal constant fraction will optimize the ranging precision.By the Monte Carlo simulation method, the underwater target echoes in different distances under different absorption coefficient and scattering coefficient are simulated. Then, the calculation results are processed by the proposed adaptive filter algorithm and the effects are evaluated by signal-to-backscatter ratio(SBR). The variation trend of detection signal with the water parameter changes is obtained. The universality of the proposed water backscattering filter is verified in theory.According to the above theoretical results, the underwater Lidar system prototype based on the adaptive backscattering filtering technique is developed and assembled. The prototype reaches design requirements of miniaturization, automation and intelligentialize, which meet the requirements requirements of unmanned underwater vehicle. Through 0.4Mpa hydrapress test for three hours, the water tightness and press-resistance are proved, the system satisfies the 40 meters underwater work demands.Finally, the system is debugged and calibrated. Then we have successively completed the underwater detection experiment in basin and the confirmatory test in the natural water. In the basin experiment, the underwater target can be detected from up to 26 meters away. The optimal constant fraction is set to 60%, and ranging accuracy is 0.3 m under this condition, which is better than the one in traditional method. The SBR values of processed signals by VFF-RLS method remain above 15 dB. The increments of SBR from 11 m to 26 m target processed signals are proportion to target distance. The increment values changes between 18.9 d B and 43.1 dB. The proposed method in this paper can effectively restrain water backscattering signals in the detection echoes of different distances. Especially in weak signal from far distance target, the target contrast enhance obviously. The above detection and working performance are also verified in the natural water environment. The prototype has the ability to detect the underwater target within 20 meters and measure the range in natural waters.