基于数码摄像的蓝藻识别研究及工程实现
2016-10-21王丽霞马美英许媛等
王丽霞 马美英 许媛等
摘要针对近年来我国内陆水体富营养化程度日益严重,导致蓝藻水华大规模爆发的情况,为了简单快速识别蓝藻,结合当前数码摄像技术不断提高的实际,提出并实现了利用高分辨率、高质量的图片对蓝藻的关键特征进行识别的方法。该方法从空间域和频域这两个领域对蓝藻图片进行分析,首先引入色彩空间原理,根据预先所设的蓝藻特征值门限处理图片,提取蓝藻在空间域上的颜色特征;然后利用傅里叶变换把空间域信号变换到频域,并通过带通滤波及高通滤波技术分析蓝藻图片,得到蓝藻在频域上的纹理特征;最后结合这两个不同领域各自的结果交叉识别图片中的蓝藻。试验结果表明,该方法能够正确得到蓝藻图片的颜色特征和纹理特征,通过两个特征能很好地识别蓝藻,取得了较好的效果。
关键词蓝藻识别;颜色特征;傅里叶变换;纹理特征
中图分类号S126文献标识码A文章编号0517-6611(2016)04-304-05
Research and Engineering Realization of Cyanobacteria Identification Based on Digital Photography
WANG Lixia, MA Meiying, XU Yuan, Shao Yubin*(College of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, Yunnan 650000)
AbstractIn recent years, Chinas inland water eutrophication is becoming more and more serious, which results in algal outbreak in large scale. With the improvement of digital photography technology, a method was proposed and realized to identify cyanobacteria better and faster, which makes use of the high resolution and high quality images to identify the key features of cyanobacteria. This method analyzed the images in spatial domain and frequency domain. A spatial color principle was introduced firstly, images were handled according to predetermined Double Eigenvalue Threshold of cyanobacteria to get the color characteristics of spatial domain, and then the signal of spatial domain was transformed to frequency domain by Fourier transform. At the same time, cyanobacteria images were analyzed by bandpass filtering and highpass filtering technology, so as to obtain the texture characteristics on the frequency domain. Finally, the cyanobacteria in the images were identified by the results of the two domains. The experimental results showed that this method could get the color characteristics and texture characteristics from cyanobacteria images accurately, and identify cyanobacteria better by the two characteristics, and the result was satisfactory.
Key wordsCyanobacteria identification; Color characteristics; Fourier transform; Texture characteristics
根據历年湖水检测的数据,国内2/3以上的水质由于不同程度的污染呈现严重的富营养化状态[1],水体富营养化导致蓝藻大面积滋生,最终形成蓝藻水华现象,这也是水体恶化的突出表现[2]。水华现象的形成从根本上来说取决于蓝藻的生物学特性,同时还受污染水体恶化程度、富营养化元素浓度、水体流速及环境变化等影响,严重的水质污染已使蓝藻水华几乎连年爆发,其对大自然的危害程度非常大。
我国湖泊河流众多,水质监测和治理技术通过多年的积累和发展已经逐渐形成了一套行之有效的体系,但仍然不够精确地识别蓝藻。……
