Underwater Image Acquisition And Real-time Processing Under Illumination By Light Sources With Different Wavelengths

With the rapid development of the ocean exploration and underwater targetidentification, the underwater imaging technology shows wide application prospect, suchas the observation of fish and seafloor, salvage and rescue. While the effect of absorptionand scattering make it difficult to obtain clear image underwater. So far, the underwaterimaging including synchronous scanning and range gating imaging technology to a largeextent eliminates the noise pollution caused by the backscattering, but requires theilluminators and image capture devices with high performance. A new practical imagingand processing technology with simple hardware structure and less energy consumptionwould be more feasible to improve the quality and resolution of the imaging target for thecomplex underwater conditions.After the characteristics of light attenuation underwater were analyzed in details, anavailable underwater illumination scheme and a new processing technology were proposedhere. Then the light sources with different wavelengths were employed for illuminationexperiment. The different water environment was simulated by adding the blue-green inkand milk for the absorption and scattering effects respectively. That adopting different lightsources illuminate under different water environment can overcome the effects ofabsorption and scattering to ensure the image definition. From the illumination experimentabove learnt that blue-green light has better transmission to overcome the water absorptionand the red light with longer wavelength exhibits clearer illumination to alleviate thescattering influence, an optical processing method was applied in dynamic video capture toincrease the processing speed with low energy consumption. The optical processingmethod can remove the scattering noise with the differential image grey-scale value of thered light and green light collected to obtain target image information. We demonstrated themethod on processing static images in Windows systems and achieve a high performance.Comparing with the USM sharpening algorithms, it has the advantage of simplercomputation and good real-time performance, which was more feasible for the embeddedsystems of low energy consumption and limited hardware equipment.On the other hand, the image processing technology was also applied to the videocapture in Android system. This paper design an external LED strobe applied in Androidsystem with high power lighting (30w) to acquire the static image. With the purpose ofenergy saving for taking picture, Android system sent the trigger signal of LED lightsource by Bluetooth to achieve the synchronization of strobe illumination and image acquisition. At the same time, through the application programming interface of cameramodule realized the real-time video processing for enhancing image definition andresolution.