側掃聲納影像之能量補償修正研究

Abstract

側掃聲納影像的成像強度為相對性並非絕對，同一處的海床，其接收到的回散射訊號強度，會受到聲波於水中傳遞的距離、聲波頻率、聲波發射角度與聲波入射海床角度等的影響，而有不同的反應，且在海床影像成圖時，常會在拖魚正下方周遭的海床看見不規則的異常聲波能量強值，此強值的形成不僅與上述因素有關，還增加了聲納裝設角度的影響。若要將此強值消除，必須多方考量各種因素造成的反應，不是經由簡單的數值運算如平均、中位數等方法可以消除的。 本研究使用兩筆新竹外海和一筆苗栗外海的野外實測側掃聲納資料，將蒐集的原始聲波資料數值化後，分別採用Stuart Anstee(2001)年和張逸中等人(2009)提出的修正辦法，以數值計算的方式，期望能將側掃影像能量不均的現象消除。 不管是Stuart(2001)或張(2009)等人的修正辦法，都有提出海床為平坦且拖魚離海床高度不變的假設前提，此假設是為了使資料中每筆掃描線的拖魚離底高度固定，在後續修正時，才不會將水層雜訊也納入計算，但由於本研究使用的是實際探測的外業資料，拖魚沒有讓它始終保持在離海床相類似的高度，計算中一直有誤差產生，經過多次嘗試後，本研究發現，針對Stuart(2001)和張(2009)等人的能量補償方法，可以分別使用縮減選用資料的範圍以及將具有相同拖魚離底高度的掃描線提出做計算的方式，將各條掃描線海床深度不同所產生的計算誤差消除。 將新竹外海的其中一筆資料以兩種方法分別修正後，結果顯示，張(2009)等人的修正辦法可以達到較好的聲波能量補償效果，影像不僅在橫向能量上的均衡，在縱向也能達到能量上的均衡，並且在進行補償能量的數值運算時，也較易於Stuart(2001)的計算方法。 本研究修改了張(2009)等人在縱向能量均衡的修正方法，以全部掃描線的總聲波能量做平均後發現，此方法可避免因為資料量的變化，導致要隨時調整計算的掃描線數目，才能獲得較好能量均衡效果的狀況發生，也不會因為低通的效果，導致影像強度值過低或失真的現象，是一個較優於原先張(2009)等人所提出的前後各10條掃描線作計算的方法。

The imaging intensity of the sidescan sonar is not absolute but relative. In the same seabed, the intensity of the backscattered signal will be affected by the distance of acoustic wave transmitted in water, the acoustic frequency, the angle of the acoustic emission, and the angle of the seabed into which the acoustic waves are incident, etc. And in the seabed image mapping, can often see anomaly acoustic wave energy value in the seabed directly below the towfish, the formation of this strong value not only with the above factors, but also increased the angle of the sonar installation influences. To eliminate this strong value, we must take into account the various factors caused by the reaction, not by simple numerical operations such as average, median and other methods can be eliminated. In this study, two field samples of Hsinchu offshore and Miaoli offshore were used to digitize the original sonic data, and the modified method was proposed by Stuart Anstee (2001) and Zhang Yi zhong et al. (2009), In a numerical way, it is desirable to be able to eliminate the phenomenon of side-scan image energy uneven. Regardless of the modification proposed by Stuart (2001) or Zhang et al. (2009), there is a hypothesis that the seabed is flat and the height of the towfish is not changed from the seabed. This hypothesis is intended to make every scan line have the same towfish height and will not be included in the calculation of water layer noise when subsequent corrections are made. However, since this study uses field data from the actual survey, the towfish did not keep it at a similar level to the seabed, there has been error in the calculation. After several attempts, this study found that for Stuart (2001) and Zhang et al. (2009) energy compensation method, it is possible to eliminate the calculation error caused by the difference of the seabed depth of each scanning line by using the range of reducing the selection data and presenting the scanning lines having the same height of the towfish from the bottom. After one of the data in Hsinchu's offshore area was revised in two ways, the results show that Zhang et al. (2009) can achieve better sound energy compensation effect, the image is not only in the horizontal energy balance, but also in the vertical energy balance, and it is also easier than Stuart’s way when performing the numerical calculation of the compensation energy. In this study, we modified Zhang et al. (2009) vertical energy balance correction method to averaging the total acoustic energy of all scanning lines. This method can avoid the need to adjust the number of scanning lines at any time because of the change of data amount, and also will not result in the phenomenon of the image intensity value being too low or distorted because of the low-pass effect. Is a better than the original Zhang et al. (2009), who made before and after the 10 scan lines for the calculation method.