基于灰度分布信息的鲟鱼声呐图像分析研究

为证实灰度信息在声呐图像中的重要价值,并获得有利于声呐图像目标分割的最佳增益值,研究了不同声呐增益值下,鲟鱼声呐图像灰度分布的形态差异,并基于灰度直方图分布信息的最大熵阈值分割算法对俄罗斯鲟（ Acipenser gueldenstaedti Brandt ）目标进行了分割,根据分割结果得到了最佳声呐增益值,即Gain＝15 dB。充分说明了在较优的声呐增益值下获取的声呐图像,即使目标形状发生严重变形时,利用其灰度直方图分布信息也能进行很好的目标分割,从而为解决鲟鱼声呐图像目标识别以及跟踪提供参考依据。     

一种基于前视声呐的养殖网箱内鱼群数量评估方法

为准确、高效评估养殖网箱场内鱼群数量，科学指导网箱养殖，提出了一种基于前视声呐的养殖网箱内鱼群数量统计方法。方法应用时，以声呐图像处理技术为基础，首先以目标网箱内养殖鱼类在消声水池实验室的声学标定结果为依据，对前视声呐测量声图进行单个鱼体的声学图像阈值划分；其次凭借云台控制声呐旋转，对围栏区域进行遍历探测，随后将声学标定结果确定的阈值作为划分依据，将不同角度所测声学图像中鱼类进行划分、统计，估算得到整个养殖网箱内的鱼类数量；最后结合探测角度以及位置信息，得到网箱养殖区域内的鱼类空间分布特征，用于指导渔民进行后续的网箱养殖。此方法在舟山青浜岛大黄鱼养殖网箱中得以应用。结果显示：该方法估计得到的养殖网箱内大黄鱼数量与人工评估结果基本一致，经多次计算，平均相对误差仅为2.45%,且实现过程更加高效、便捷。本研究能够有效为养殖网箱提供鱼类数量信息，为亟待进行鱼类数量监管的网箱养殖产业提供科学、可靠的数量统计方法，具有广泛的应用场景。     

浅析唐山市海洋牧场建设的必要性

正海洋牧场是指人们利用自然海洋环境,通过建造适宜的渔业设施、投放人工培育的增殖苗种和采取鱼群控制技术等措施将目标经济生物聚集起来,形成大型人工渔场,以提高渔业产量和保护海洋生态系统,实现渔业可持续发展。世界海洋大国都在积极发展海洋牧场,其中美国、日本、韩国的海洋牧场建设处于世界先进水     

气泡幕对许氏平鲉的阻拦效果研究

气泡幕可作为海洋牧场鱼群控制的技术手段之一。本研究通过改变气泡幕在水槽中的相对位置和试验许氏平鲉(Sebastes schlegeli)的密度来观察许氏平鲉的行为反应,运用了观察法、对照试验法和控制变量法探究了气泡幕对不同密度下的许氏平鲉的阻拦效果。结果显示:1)试验鱼20尾时,气泡幕的位置分别设置在矩形水槽长的1/4、1/2和3/4处(鱼的密度为12.52、6.26、4.17尾/m2),阻拦率分别达到了98.2%、88.4%和92.9%,阻拦效果明显。2)气泡幕设置在水槽长度方向的1/2处,试验鱼分别为5尾、10尾、15尾和20尾时,气泡幕阻拦率分别达到了100%、85.7%、100%和88.4%(鱼群密度分别为1.56、3.12、4.69、6.26尾/m2),阻拦效果良好。3)由于阻拦率较高,试验鱼群密度和阻拦效率没有明显的关系。本研究结果可为我国投放许氏平鲉的海洋牧场鱼群控制技术以及取水口拦鱼技术提供参考。     

面向海洋牧场的水下机器人强度校核

随着海洋牧场的建设与开发,需要水下机器人代替人在水中执行各种任务。团队设计一款面向海洋牧场的水下机器人,可以进行水下捕捞及水下环境的监测等。迎流耐压壳体是水下机器人的重要承压部分,目前尚无对其结构的强度和稳定性进行评估的相关规范方法。选取水下机器人的耐压壳体这个关键部件,利用ANSYS Workbench对其进行强度分析,结果表明:水下机器人的耐压强度满足设计要求,使用安全可靠。     

基于Ransac算法的多波束点云人工鱼礁提取方法

投放人工鱼礁作为海洋牧场建设的重要环节，其投放质量评估工作日益受到重视。为实现大范围的人工鱼礁三维信息的自动识别与提取，提升人工鱼礁量化评估的质量，提出了一种基于Ransac算法的多波束点云人工鱼礁提取方法。该方法首先利用多波束声呐获取点云，在对点云进行分割、去噪等预处理后，利用Ransac算法将点云分为内外点集，内点集近似拟合为海底平面，外点集为人工鱼礁。在浙江嵊泗马鞍列岛某礁区进行了相关试验，实现了整体和单体鱼礁的提取，提取正确度达94.79%,完整度达91%,能够较好地提取鱼礁的三维形态、位置、尺寸、高度、空方等信息。研究表明，本方法可以大范围地应用于各型人工鱼礁目标投放效果的量化评估。     

音响驯化对许氏平鲉有效作用范围的研究

音响驯化型海洋牧场的建设对于渔业资源的增殖、海洋环境保护、渔业可持续发展意义重大。研究音响驯化有效作用范围，可以为海洋牧场人工鱼礁的投放选址提供科学合理的参考依据。本研究参照T/SCSF 0009—2021《海洋牧场鱼类音响驯化技术规程》设置许氏平鲉(Sebastes schlegelii)音响驯化放声参数，在海洋牧场示范区通过水下播放300 Hz正弦波模拟野外音响驯化，通过测算建立水下噪声衰减模型。结果显示：水下1 m、5 m、10 m的水平衰减公式分别为T_TL=1.73×10 log(r)、T_TL=1.58×10 log(r)、T_TL=1.70×10 log(r),水下噪声衰减为球面衰减。在海洋牧场示范区内对许氏平鲉选用参数为300 Hz、 155 dB的正弦波进行音响驯化的有效半径为500 m,音响驯化的最佳半径为250 m。     

粤东柘林湾海洋牧场景观结构与格局的分析研究

海洋牧场与休闲渔业融合发展是我国海洋牧场发展方向之一,现阶段海洋牧场休闲产业规模较小,类型较为单一,主要以游钓和潜水休闲为主,其景观观赏性功能有待深入挖掘和开发。海洋牧场景观格局研究主要针对海洋牧场景观资源的形成、分布及其在景观空间上排列和布局等方面,发现并了解其生态系统或空间属性变异程度表观现象。本研究以粤东柘林湾海洋牧场示范区为案例,采用景观格局分析法对景观要素特征指数及景观异质性指数进行计算,研究表明,粤东柘林湾海洋牧场示范区自然景观为景观基质,半自然景观斑块面积最多,故破碎度指数最高,受人类活动影响最大,对生物多样性保护及景观生境维护应加强;文化景观与整体景观分离度较大,导致整体景观连通性较差,影响景观效果;滨海景观斑块较少,丰富度不高,但根据调查显示周边旅游资源丰富,因此缺乏与粤东柘林湾海洋牧场结合开发利用。总之,本研究通过解析海洋牧场景观格局空间内不同景观的分布状况及相互之间的作用和影响,为下一步进行景观生态规划提供数据支持,并为开发和利用海洋牧场休闲渔业与生态景观可持续化发展,综合提升海洋牧场的生态、经济和社会价值提供支撑。     

国内外海洋牧场发展历程与定义分类概述

海洋牧场是国内外长期广泛应用的海洋渔业发展举措。本文对国内外海洋牧场发展情况以及已有定义、分类情况进行了总结和分析。国外普遍将渔业资源增殖等同海洋牧场,从20世纪40年代以来人工鱼礁已从较单纯的聚鱼功能,拓展为海洋牧场生境保护和修复功能,目标是渔业资源增殖和利用。东北亚国家如日本、韩国建设人工鱼礁的目标是大规模增殖、聚集和捕获渔业资源;欧美诸多国家建设人工鱼礁主要是为了增殖保护渔业资源和开发休闲渔业。国外以渔业资源增殖为主旨的海洋牧场建设成败皆有,成功的方面仍值得中国借鉴。国外对海洋牧场定义和分类的数量较少,定义的提出和使用相对严谨。多年来,中国高度重视海洋牧场建设,将其作为水生生物资源养护、渔民转产转业和海洋生态文明建设的重要手段加以扶持。海洋牧场在建设规模、产出效果、技术水平、发展机制等方面取得了巨大进步。但由于起步较晚,产业和科技总体上仍处于初级阶段,存在一系列问题亟待解决。在20世纪60年代至70年代,国内学者提出海洋牧场和海洋农牧化概念,并将二者进行了区分,其后诸多学者也将渔业资源增殖等同海洋牧场。近年来国内海洋牧场发展迅速,但海洋牧场定义不够规范,数量繁多、含义空泛、严谨性不足。最后,本文就海洋牧场发展趋势、发展理念、目标定位、建设内容、研究支撑等方面进行了总结和展望,认为实事求是并理性把握好现阶段海洋牧场的发展理念、定位、内容和目标,加强对农业农村部现有海洋牧场水产行业标准的宣传尤其是定义和分类的宣传,引导海洋牧场研究和建设工作有序开展,对于切实有效支撑海洋牧场健康发展十分重要。     

基于图像处理的鱼群运动监测方法研究

鱼类运动行为的观察能够为鱼类健康监控提供直观信息,而通过人工标定的方式监测鱼群运动行为耗时长、效率低。文章针对鱼类运动行为的监测问题,提出一种基于图像处理技术的罗非鱼运动监测方法。首先利用计算机、CCD高清摄像机获取鱼群运动视频,再对图像进行滤波去噪、灰度等处理;通过Ostu阈值分割法改进Canny边缘检测算法提取鱼群的边缘轮廓;在建立鱼群运动模型的基础上结合目标关联匹配算法,实现罗非鱼运动行为的跟踪和监测。结果显示鱼群的个体检出率为98.96%,轨迹完整度为97%。提出的算法比卡尔曼滤波的轨迹跟踪监测效果略有提升,能够较好地完成鱼群的运动跟踪和动态监测。     

Automated acoustic method for counting and sizing farmed fish during transfer using DIDSON

Counting and sizing large farmed fish such as tuna is often performed during their transfer from one net cage to another. Dual-frequency identification sonar (DIDSON) provides an automated fish counting and sizing tool. However, its counter and sizer are not suitable for measuring farmed fish because of net movements due to currents and subsequent frequent image breakups. This paper presents a fully automated acoustic method to count and size farmed fish during fish transfer by using DIDSON imaging. The background is subtracted from the image after being stabilized by an image phase-only correlation method. The segmentation of the fish is obtained by tracing the edges with a contour tracing method. To prevent recounting the same fish, a Kalman filter algorithm was designed and adapted to predict fish movements. Automated counting was performed by analyzing the spatiotemporal trajectory of the track. The separated fish images were searched for and body length was obtained by summing down the centerline segments from the head to the tail of the fish. The proposed system was verified using farmed yellowtail, Seriola quinqueradiata (mean total length 83.1 cm) to obtain a sizing error of mean total length within 2.4 cm.     

基于高精度声学图像的海洋牧场海床类型识别

人工鱼礁作为海洋牧场建设的基础生态工程,在海洋生态环境修复和渔业资源增殖方面起着重要作用。随着人工鱼礁的大量投放,海洋牧场海床逐步转变为高度异质性的底栖生境,底栖生态变化会直接影响到底栖及底层生物的栖息与繁衍。针对传统调查方法在海洋牧场中的局限性,聚焦天津海洋牧场,基于侧扫声纳后散射数据和多波束测深数据,从侧扫图像亮度、图像纹理及地形起伏度等3个维度,探索了天津海洋牧场海床类型识别、分类和面积量化的新方法。研究表明,天津海洋牧场共发现人工鱼礁、牡蛎类、淤泥类、蛤仔泥混类、牡蛎泥混类及碎贝泥混类等6种海床类型,基于声学图像特征的海床类型识别方法准确率达90%,可满足海洋牧场底栖生态研究的需求;通过对比分析,天津海洋牧场本底淤泥类海床消退为58.19%,贝类海床增至7.96%,贝类资源增殖效果明显;人工鱼礁周边新生牡蛎礁0.246 km2,并伴有0.303 km2的牡蛎泥混海床待发育为牡蛎礁,海洋牧场建设对牡蛎礁具有重要的修复作用。本研究探索了适宜于海洋牧场的海床声学分类方法,探讨了声学技术对传统底栖调查方法的补充与完善作用,明确了海床类型与声学图像特征的紧密联系,验证了海洋牧场对牡蛎礁的修复作用,成果可有效应用于近岸海域生境修复、贝类资源量精准评估及牡蛎礁演化机理等领域的研究。     

Integration of sonar and optical camera images using deep neural network for fish monitoring

Fish monitoring in aquaculture farms is indispensable for managing the growth and health status of fish resources. However, it is unrealistic to expect humans to be able to perform monitoring at night, when standard optical cameras are generally inapplicable. Although sonar systems can be used at night, their practical applications are limited by their monochrome, low-quality images. In this paper, we describe a realistic image generation system that uses sonar and camera images recorded at night. The proposed approach is based on conditional generative adversarial networks, which learn the image-to-image translation between sonar and optical images. We tested the system in a fish tank containing thousands of sardines (Sardinops melanostictus). Images were simultaneously recorded using high-precision imaging sonar and an underwater camera. Experimental results show that the proposed model successfully generates realistic daytime images from sonar and night camera images. Our system enables nighttime monitoring using sonar and an optical camera, leading to more efficient fish farming and environmental surveillance.     

Acoustic approaches to remote species identification: a review

Noninvasive species identification remains a long-term goal of fishers, researchers, and resource managers who use sound to locate, map, and count aquatic organisms. Since the first biological applications of underwater acoustics, four approaches have been used singly or in combination to survey marine and freshwater environments: passive sonar; prior knowledge and direct sampling; echo statistics from high-frequency measures; and matching models to low-frequency measures. Echo amplitudes or targets measured using any sonar equipment are variable signals. Variability in reflected sound is influenced by physical factors associated with the transmission of sound through a compressible fluid, and by biological factors associated with the location, reflective properties, and behaviour of a target. The current trend in acoustic target identification is to increase the amount of information collected through increases in frequency bandwidth or in the number of acoustic beams. Exclusive use of acoustics to identify aquatic organisms reliably will require a set of statistical metrics that discriminate among a wide range of similar body types at any packing density, and incorporation of these algorithms in routine data processing.     

The application of multibeam sonar technology for quantitative estimates of fish density in shallow water acoustic surveys

The paper describes the main drawbacks in the application of conventional acoustics in shallow waters, and reviews the advantages and limitations that existing multibeam sonar present in these ecosystems. New techniques and methods for adapting multibeam sonar to shallow waters are proposed and discussed. A method for analysing acoustic data from shallow waters through image analysis process is presented and some examples are considered. The results show that scattered fish can be observed individually and counted, and that schools are described in their morphology and behaviour. From these results an ‘ideal’ acoustic device is defined: a sonar operating at more than 400 kHz with a coverage of at least 120° in one direction and, depending on the needs of the user, 15° or 1° (which can be modified easily) in the perpendicular plane. The beam opening–angle is 0.5° in the centre beam, increasing to 1.0° at the 60° steer–angle, giving a total of 240 beams. Multibeam sonar data could be used for several purposes in shallow waters, in particular to estimate fish density and biomass, and study spatial and temporal behaviour of fish.     

Evaluation of a stimulus-response method for distinguishing out-migrant salmonids from drifting debris for sonar counts in the Trinity River,California

A method for discriminating fish from debris in sonar counts of out-migrant salmonids was tested in the Trinity River, California. The method used induced fish movements to distinguish fish from drifting debris. Electricity and light served as stimuli and video and split-beam sonar were used to measure movements of fish (mainly juvenile chinook salmon) and debris (mainly tree leaves). Differences in fish and debris behavior were clearly observable with underwater video. Many fish darted or slowed and most fish dove, whereas debris drifted passively. Fish responded to the electric field inconsistently, and an apparent positive phototaxis was the most consistent response to stimuli. Lack of matched sonar and video observations of individual targets prevented direct testing of sonar’s ability to differentiate fish and debris in the Trinity River. However, analysis of sonar data from a similar situation in the Seton River, British Columbia, indicate that fish responses measured in the Trinity River by video were within the resolution of split-beam sonar. Split-beam position measurement error averaged ≤ 0.06 m within 5° of the acoustic axis, compared to a mean diving reaction of 0.11 m by salmonids observed with video. Proper transducer deployment, improved sonar analysis methods, and perfection of stimuli to elicit obvious and consistent fish responses are key to the success of this technique. With suitable development and validation, the stimulus-response method could become a useful tool for apportioning sonar counts among fish and debris.     

基于声学摄像仪的溯河洄游幼香鱼计数

至今主要靠目视计数来掌握香鱼(Plecoglossus altivelis)的资源量,这种方法很耗劳动力且精度低.由华盛顿大学研发的声学摄像仪DIDSON(Dual-frequency IDentification SONar)可在浑浊不清的水中摄取接近光学视频画质的图像.其附带软件虽可对体长为几十厘米的大个体鱼计数,但对于体长6 cm以下的小个体鱼,由于其回波信号弱,计数显得非常困难.本文提供了一种基于DIDSON的溯河洄游幼香鱼计数的新方法.首先从声学图像中消除静止的背景,提取出运动目标并除去噪音像素,然后用卡尔曼滤波器对剩下的目标进行跟踪,根据目标移动的特性把目标分为香鱼和垃圾2大类,在除去向下游移动的垃圾之后,可正确地计算向上游移动香鱼的数量.通过现场实测验证了上述声学计数方法的实用性和高效性.     

An evaluation of a steerable sidescan sonar for surveys of near-surface fish

A towed 330 kHz sector scanning sonar was evaluated for surveys of near-surface migratory salmonids (Oncorhynchus sp.) and other fish in the Strait of Georgia near the mouth of the Fraser River near Vancouver, BC (Canada). Detection of large near-surface fish presents a serious challenge to conventional net trawls or echo-sounder surveys due to vessel avoidance behaviour by the fish. The sonar was towed beside a research vessel on a custom-built towfish at a nominal depth of 18 m. The optimum configuration was determined to be with the sonar scanning a 30° vertical sector oriented perpendicular to the tow direction, at a ping rate of 5 Hz and a sector scan period of 5 s. Fish were detected up to a range of 100 m, limited by systemic noise and backscattered reverberation from the sea surface. Some evidence of vessel-avoidance behaviour by the fish was observed. Near surface temperature and salinity gradients are shown to induce downwards refraction of the sonar beam, resulting in biases in apparent depth and target strength of the fish. A simple model of the sonar performance in this scanning mode, including beam-pattern effects, is used to generate curves of fish target detectability and beam-pattern induced bias as a function of range and fish target strength. After correction for refraction effects, beam-pattern bias, and detectability, the survey results show reasonable agreement with net trawl and riverine escapement data.     

多波束渔用声呐作用距离预报建模及性能分析

多波束渔用声呐是海洋渔业实现高效、精准和选择性捕捞的重要工具。为了系统分析多波束渔用声呐探测性能及主要制约因素对作用距离的影响,针对圆柱形阵多波束渔用声呐,采用水声学系统仿真的方法,分别对主要为噪声和混响环境下的渔用声呐探测性能进行分析,对不同海况下噪声不同目标强度鱼群的探测性能进行比较,分析窄带系统和宽带系统降低混响的能力,并对噪声和混响混叠环境下的渔用声呐探测性能进行仿真分析;最终,从系统设计、发射方式、接收方式等多个角度提出提高渔用声呐探测性能的方法,为远距离、高分辨率多波束渔用声呐的研制提供改进的方向。