鱼雷技术
魚雷技術
어뢰기술
TORPEDO TECHNOLOGY
2013年
4期
268-271
,共4页
波导不变量%干涉条纹%Sobel锐化算子
波導不變量%榦涉條紋%Sobel銳化算子
파도불변량%간섭조문%Sobel예화산자
waveguide invariant%interference stripes%Sobel sharpening operator
波导不变量β是用来刻画由于不同简正波间相互干涉形成声场强度在距离-频率的干涉条纹图案的一个标量。通过β定义式可以利用干涉条纹图案斜率、频率来实现目标被动定位。基于此，本文提出了一种利用Sobel锐化算子来进行图案边缘检测得到干涉条纹图案斜率的方法：分割图案，确保各个小窗口中条纹斜率基本相同，利用 Sobel 锐化算子进行边缘检测得到斜率，再利用β定义式估计距离，最后综合各个小窗口的估计即可得到更加稳健的距离估计。通过对干涉条纹图案斜率进行计算可以实现目标的被动定位。计算机仿真得出距离估计良好，最大误差在10%以内。湖上试验也验证了采用该方法获得的目标距离估计最大误差在25 m以内，表明该方法简单、有效，具有良好的应用前景。
波導不變量β是用來刻畫由于不同簡正波間相互榦涉形成聲場彊度在距離-頻率的榦涉條紋圖案的一箇標量。通過β定義式可以利用榦涉條紋圖案斜率、頻率來實現目標被動定位。基于此，本文提齣瞭一種利用Sobel銳化算子來進行圖案邊緣檢測得到榦涉條紋圖案斜率的方法：分割圖案，確保各箇小窗口中條紋斜率基本相同，利用 Sobel 銳化算子進行邊緣檢測得到斜率，再利用β定義式估計距離，最後綜閤各箇小窗口的估計即可得到更加穩健的距離估計。通過對榦涉條紋圖案斜率進行計算可以實現目標的被動定位。計算機倣真得齣距離估計良好，最大誤差在10%以內。湖上試驗也驗證瞭採用該方法穫得的目標距離估計最大誤差在25 m以內，錶明該方法簡單、有效，具有良好的應用前景。
파도불변량β시용래각화유우불동간정파간상호간섭형성성장강도재거리-빈솔적간섭조문도안적일개표량。통과β정의식가이이용간섭조문도안사솔、빈솔래실현목표피동정위。기우차，본문제출료일충이용Sobel예화산자래진행도안변연검측득도간섭조문도안사솔적방법：분할도안，학보각개소창구중조문사솔기본상동，이용 Sobel 예화산자진행변연검측득도사솔，재이용β정의식고계거리，최후종합각개소창구적고계즉가득도경가은건적거리고계。통과대간섭조문도안사솔진행계산가이실현목표적피동정위。계산궤방진득출거리고계량호，최대오차재10%이내。호상시험야험증료채용해방법획득적목표거리고계최대오차재25 m이내，표명해방법간단、유효，구유량호적응용전경。
Waveguide invariant β is a scalar for characterizing the distance-frequency interference stripes, which is formed by the acoustic field due to mutual interference between different normal wave modes. Target can be passively positioned by the slope of the interference stripes and the frequency based on the definition ofβ. In this paper, a method for calculating the slope of the interference stripes is presented by taking advantage of Sobel sharpening operator to detect pattern edge. First, the pattern of the interference stripes is divided into smaller parts to keep the slope of stripes in each part approximately constant. Then, the Sobel edge detection method is employed to select the intermediate value and calculate the slope in the frequency range of the smaller part, further to obtain estimation value of distance by using the formula ofβ. Finally, the same on a different frequency range smaller pattern estimated value to the average to a more accurate estimation can be obtained by integrating the distance estimations of all smaller parts. Passive positioning of target can be realized by calculating the slope of interference stripes. Simulation result shows that the proposed me-thod possesses a good performance in distance estimation, and the maximum relative error of target distance estimation is smaller than 10%. Lake trial indicates that the maximum error of its target distance estimation is smaller than 25 m.
