构建VED-SegNet分割模型提取鱼类表型比例

为提高鱼类表型分割精度和准确度,实现鱼类表型智能监测,该研究基于深度学习算法构建了VED-SegNet模型用于鱼类表型分割和测量。该模型将cross stage partial network和GSConv结合作为编码器（VoV-GSCSP）,保持足够精度的同时降低网络结构复杂性。另一方面,该模型采用EMA(efficient multi-scale attention module with cross-spatial learning)建立强化结构,加强编码器和解码器之间的信息传递,提高模型精度,并实现了8个表型类别的输出。采用自建的鱼类表型分割数据集对VED-SegNet模型进行了测试,测量结果中鱼类各表型比例与实际测量值相接近,表型最大平均绝对和平均相对误差为0.39%、11.28%,能实现无接触式提取水产养殖中鱼类表型比例。对比其他常见语义分割模型,平均交并比mean intersection over union,mIoU和平均像素准确率mean pixel accuracy,m PA最高,分别到达了87.92%、92.83%。VED-SegNet模型在环境复杂、多鱼重叠的...     

浙江省近海渔运船转载信息提取

渔运船是从事渔获物运输的专用船舶,能够提高捕捞渔船作业效率,增加捕捞渔船的作业强度。为掌握渔运船在海上的转载情况,从而间接了解捕捞渔船作业强度,该研究提出一种基于北斗船位数据的以设定航速阈值、距离阈值和时间阈值来提取渔运船转载信息的方法。如果渔运船和捕捞渔船距离小于50 m,且期间有持续3条以上的船舶监控系统(Vessel Monitoring Systems,VMS)记录,则认为可能发生了1次转载,并记录下相遇的时长、船名、空间位置。以浙江省为例,利用该方法从2018年浙江省的北斗船位数据中提取渔运船的海上转载信息,并进行统计分析。结果表明,有转载记录的渔运船808条,参与转载的捕捞渔船3 548条,共转载28 916次。渔运船停船转载占比21.0%,以1～1.4 m/s低速航行的作业状态转载占比53.7%,转载时长小于12.5 min的占比81.3%,同时得到渔运船转载的热点分布,转载累积时长最长的空间网格为122.5°E～123°E,31.5°N～32°N,转载累积时长187 h,其次为122°E～122.5°E,28°N～28.5°N,转载累积时长150h。通过分析渔运船海上转载位置和转载累积时长的空间分布情况可掌握捕捞渔船作业的时空变化特点,为渔业限额捕捞精细化管理提供依据。     

权重约束AdaBoost鱼眼识别及改进Hough圆变换瞳孔智能测量

针对传统鱼眼瞳孔直径测量方法耗时、耗力,且数据主观性强的问题,该文提出基于权重约束AdaBoost和改进Hough圆变换的鱼眼瞳孔直径智能测量方法。首先,利用工业相机采集实验板上的鱼图像,从正负鱼眼图像样本中训练出基于权重约束AdaBoost算法的鱼眼分类器;然后,采用该分类器对试验图像进行检测,将检测到的鱼眼局部图从整体图中分离出来;最后,采用改进的Hough圆变换检测出鱼眼的瞳孔,并计算得到瞳孔直径。对100条金鲳鱼进行试验,鱼眼分类精度达97.1%,瞳孔正确检测率达94.2%,相比改进前分别提升了1.7个百分点和10.5个百分点,与人工测量瞳孔直径值的平均偏差为6.5%,比改进前低了5.9个百分点,总的平均测量时间为324.371 ms,比改进前减少了10.707 ms。试验证明:该文提出的方法能够精确、实时、自动地测量出鱼眼瞳孔的直径,有效避免了传统测量方式的复杂性和测量数据的主观性,可为鱼体生长状况评估、良种选育提供重要参考。     

自然水体中超声波标记鱼游动轨迹精密确定算法

针对鱼类关键生境位置确定的应用需求,该文提出了一套适用于自然水体的超声波标记鱼定位算法,解决了标记鱼定位以及存在粗差观测值,即水听器记录的超声波信号接收时间存在错误情况下算法的抗干扰性。宜昌黄柏河的实测结果表明,基于现有1 ms级精度的水听器,可在自然水体中获得2.15 m精度的信号标记鱼三维游动轨迹。如因气泡、遮挡等因素对水听器观测数据引入粗差,当粗差量级在10 m以上,该方法可接近100%探测出是否存在粗差。当粗差观测值在3个以内时,该方法的探测成功率可达84.3%以上,3个以上时粗差探测成功率明显下降,5个及以上,即粗差观测值个数占观测值总数的比例大于31.25%时,基本只能探测出观测数据中存在粗差而无法有效确定粗差。该研究可为渔业增殖、鱼类栖息地保护、鱼类洄游通道等研究提供参考。     

鱼粉品质检测电子鼻传感器阵列的多特征数据融合优化

为了提高基于仿生嗅觉的鱼粉品质检测装置的鉴别能力,该文利用自主设计的仿生嗅觉鱼粉品质检测装置,提取鱼粉样本的响应特征信息,对其传感器阵列进行多特征数据融合优化。依据各传感器对样本的响应曲线,提取传感器特征值(10×6个)构成原始特征矩阵,后对传感器阵列特征值进行归一化处理,以紧凑性作为评价特征选择方法合理性的标准,采用了3种单特征排序方法(MIC、χ^2、F-test),3种多特征排序方法(RF、LR、SVM),4种特征递减消除方法(RFRFE、SVMRFE、DTRFE、LRRFE)对不同品质的鱼粉进行分类准确率检验,得到基于随机森林的特征递减消除算法(RFRFE)的紧凑性最好,此时最佳的分类准确率为98.3%,特征数目为33个。优化后的传感器阵列特征发生了明显的变化,传感器阵列由原来的10个变为了8个,去掉了传感器TGS2620和传感器TGS2600,特征值也减少了45%。为了避免选择偏差,采用了10折交叉验证方法,再次得到了RFRFE算法具有更佳的紧凑性。该特征选择方法为利用仿生嗅觉技术鉴别其他动物源性原料样本的特征优化提供了新的方法和参考。     

水蜜桃果汁果冻的加工工艺

以水蜜桃为主要原料,采用果冻加工工艺制备果汁果冻。结果表明,采用50%水蜜桃汁、6%白砂糖、0.1%甜味剂蛋白糖FT-50、0.1%的琼脂与0.2%CMC-Na的工艺参数,可制出色泽均匀、组织状态良好、口感细腻、酸甜适宜的水蜜桃果汁果冻。果汁果冻的杀菌方式采取超高温瞬时灭菌,该方法对于VC的保持率达到了65.10%。     

Tracking change in a human‐dominated landscape: developing conservation guidelines using freshwater fish

• 1. To conserve biodiversity in a human‐dominated landscape, a science‐based inventory and monitoring plan is needed that quantifies existing resources, isolates drivers that maintain natural communities, determines harmful stressors, and links ecological drivers and human stressors. A tactical approach is proposed for conservation planning using freshwater fish at the Cape Cod National Seashore.
• 2. Freshwater fish are well studied and occur across environmental gradients. The lentic systems at the national park are relatively pristine yet are enveloped by a region of high population density. Using fish community data, three steps were taken for tracking anthropogenic impacts in a human dominated landscape. First, fish and potential drivers were sampled intensively along a gradient to determine which fish metrics reflect natural communities and which abiotic and biotic factors structure them. Second, emerging and existing regional human threats were identified. Third, these human threats were linked to the potential drivers that maintain natural communities to identify the most informative metrics to monitor and track change.
• 3. Fish communities, water quality, habitat, and food resources were sampled concurrently in 18 ponds in 1999 and 2000. Three common fish species explained 98% of variation in numbers across systems. Based on ecological relationships, pH, depth, vegetation, prey, and community complexity were determined to maintain biodiversity of freshwater fish communities.
• 4. The primary human threats here included: development‐related, land‐use changes; non‐point source pollution; eutrophication from septic systems; and introduced species that are a byproduct of high human visitation. These are common threats in many rapidly urbanized areas and are likely to have relevance to many sites.
• 5. To track the impact of emerging threats to freshwater ponds related to increased human population, monitoring changes in water quality, vegetated habitat, fish diversity, and trophic interactions are recommended.

Copyright © 2008 John Wiley & Sons, Ltd.     

Apparent digestibility coefficient of protein and amino acids of acid, biological and enzymatic silage for Nile tilapia (Oreochromis niloticus)

Using the fish silage to partially replace proteic feedstuff in aquafeeds is an alternative to mitigate sanitary and environmental problems caused by the lack of adequate destination for fisheries residues. It would also lower feed costs, consequently improving fish culture profitability. However, using fish silages in aquafeeds depends on determination of its apparent digestibility coefficients (ADC). This work aimed to determining the ADC of crude protein and amino acids of acid silage (AS), biological silage (BS) and enzymatic silage (ES) for juvenile Nile tilapia (94.5 ± 12.7 g). The ADC_CP was: 92.0%, 89.1% and 93.7% for AS, BS and SE respectively. The average ADC of amino acids was: 91.8%, 90.8% and 94.6% for AS, BS and ES respectively. Results encourage the use of AS, BS and ES to partially replace protein sources in balanced diets for neotropical fish.     

Determination of apparent digestibility coefficient in fish by stable carbon isotopes

Estimation of the apparent digestibility coefficient ADC(%) of C₃ and C₄ plants in experimental diets for fingerlings of tambaqui (Colossoma macropomum, Cuvier, 1818) was calculated by applying chromic oxide (Cr₂O₃) external marker methodology and by a proposed mathematical expression based on the isotopic composition (δ¹³C). A total of 240 tambaqui fingerlings each weighing ±48.2 g and measuring ±9.8 cm were maintained in eight 500‐L aquariums specially designed for faeces collection. The ADC(%) of the C₃ and C₄ diets did not differ significantly between the two methods, producing results of 75.6%; 76.2% and 74.4%; 72.8%, respectively. The ADC(%) results obtained by isotopic method presented less variation than by chromic oxide. The proposed mathematical expression for calculating the ADC(%) based on δ¹³C values offers an alternative methodology, which can reduce errors and diminish the effort required to collect biological material. However, it is important to note that this method is limited to analysis of diets or food items with distinct isotopic signals.     

Blood and haemoglobin meal as protein sources in diets for gilthead sea bream (Sparus aurata): effects on growth, nutritive efficiency and fillet sensory differences

Two parallel experiments were conducted to evaluate the effect of partial substitution of fish meal by two different animal protein sources, blood meal (B) and haemoglobin meal (H) at 0, 50 and 100 g kg⁻¹ of level inclusion in diets for gilthead sea bream, considering a long feeding period for juveniles (Trial 1) and a short feeding period (Trial 2) for on-growing fish. In Trial 1, 33 g juveniles were fed for 242 days and the fish fed with 5% and 10% of haemoglobin dietary inclusion obtained less growth, although feed efficiency, protein efficiency ratio and muscle composition were similar in all diets. In Trial 2, 179 g initial weight fish were fed for 164 days and growth of fish fed H10 showed the lowest growth, although nutrient efficiency and muscle composition were not affected significantly. The results of these experiments demonstrated that blood meal can substitute fish meal (up to 10%) with no effect on performance, but may lead to sensory differences compared with fish fed diet 0, while the inclusion of 5% blood meal had no effect on growth or sensory characteristics. Fish fed 10% haemoglobin inclusion had the poorest growth values.