摘要
针对目前检测方法特征单一、样本数量少和鲁棒性低等问题,提出了一种基于多特征融合与机器学习的鱼类摄食行为的检测方法:利用图像处理技术提取鱼群摄食图像的颜色、形状和纹理特征,并对其进行归一化和特征融合处理,通过构建3层的BP神经网络对鱼群摄食行为进行检测。与SVM和KNN检测效果进行对比,BP神经网络的效果最好,精度可达97.1%。与传统的基于单一纹理特征方法相比,在保证时效性和增强鲁棒性的同时,准确率提高了4.1%。
Aiming at the problems of single feature, less data and low robustness of current detection methods for the fish feeding behavior, a detection method was proposed based on the multi-feature fusion and the machine learning, by using image processing technology to extract the color, shape and texture features of the fish feeding images, which were processed by normalization and feature fusion. The fish feeding behavior is checked by constructing a 3-layer BP neural network. The results show that compared with SVM and KNN, the BP neural network has the best effect, and the accuracy can reach 97.1%. Compared with the traditional method based on the single texture feature, the accuracy is improved by 4.1% while guaranteeing timeliness and enhancing robustness.
引文
[1]农业部渔业渔政管理局.2017中国渔业统计年鉴[M].北京:中国农业出版社,2017.
[2]胡金城,于学权,辛乃宏,等.工厂化循环水养殖研究现状及应用前景[J].中国水产,2017(6):94-97.
[3]穆春华,范良忠,刘鹰.基于计算机视觉的循环水养殖系统残饵识别研究[J].渔业现代化,2015,42(2):33-37.
[4]VASSILIS M P,ALEXIOS G,MAROUDIO K.Sub-second analysis of fish behavior using a novel computer-vision system[J].Aquacultural Engineering,2014,62:36-41.
[5]TORISAWA S,KADOTA M,KONEYAMA K,et al.Adigital stereo-video camera system for three dimensional monitoring of free-swimming Pacific bluefin tuna,Thunnus orientalis,cultured in a net cage[J].Aquat Living Resour,2011,24(2):107-112.
[6]CUI Z,WU J F,YU H,et al.A review of the application of computer vision technology in aquaculture[J].Marine Science Bulletin,2018,20(1):53-66.
[7]LIU Z Y,LI X,FAN L Z.Measuring feeding activity of fish in RAS using computer vision[J].Aquacultural Engineering,2014,60:20-27.
[8]胡利永,魏玉艳,郑堤,等.基于机器视觉技术的智能投饵方法研究[J].热带海洋学报,2015,34(4):90-95.
[9]赵建,朱松明,叶章颖,等.循环水养殖游泳型鱼类摄食活动强度评估方法研究[J].农业机械学报,2016,47(8):288-293.
[10]陈彩文.基于计算机视觉的鱼群摄食行为分析研究[D].太原:太原理工大学,2017.
[11]陈彩文,杜永贵,周超,等.基于图像纹理特征的养殖鱼群摄食活动强度评估[J].农业工程学报,2017,33(5):232-237.
[12]郭强,杨信廷,周超,等.基于形状与纹理特征的鱼类摄食状态检测方法[J].上海海洋大学学报,2018,27(2):181-189.
[13]ZHAO J,BAO W J,ZHANG F D,et al.Assessing appetite of the swimming fish based on spontaneous collective behaviors in are circulating aquaculture system[J].Aquacultural Engineering,2017,78:196-204.
[14]ZHOU C,LIN K,XU D,et al.Near infrared computer vision and neuro-fuzzy model-based feeding decision system for fish in aquaculture[J].Computers and Electronics in Agriculture,2018,2(6):114-124.
[15]韩丁,武佩,张强,等.基于颜色矩的典型草原牧草特征提取与图像识别[J].农业工程学报,2016,32(23):168-175.
[16]LI B,XU D,LAN C.Chinese moment invariants for color constant image description[J].Chinese Journal of Electronics,2010,19(2):283-287.
[17]董红霞,郭斯羽.一种结合形状与纹理特征的植物叶片分类方法[J].计算机工程与应用,2014,50(23):185-188.
[18]SAROJA G A S,SULOCHANA C H.Texture analysis of non-uniform images using GLCM[C]//Information&Communication Technologies.IEEE,2013:1319-1322.
[19]曾志,吴财贵,唐权华,等.基于多特征融合和深度学习的商品图像分类[J].计算机工程与设计,2017(11):211-216.
[20]XIE R,WANG X,LI Y,et al.Research and application on improved BP neural network algorithm[C]//Industrial Electronics&Applications.IEEE,2010.
[21]焦李成,杨淑媛,刘芳,等.神经网络七十年:回顾与展望[J].计算机学报,2016,39(8):1697-1716.
[22]ZHANG S C,LI X L,ZONG M,et al.Learning k for kNN classification[J].ACM Transactions on Intelligent Systems&Technology,2017,8(3):43.