Research on weeds identification based on K-means feature learning
- 111 Downloads
This paper aims to overcome the unstable identification results and weak generalization ability in feature extraction based on manual design to realize the automatic weeds identification. On the basis of unsupervised feature learning identification model, K-means clustering algorithm after data preprocessing is used to realize feature learning and construct feature dictionary. Then this feature dictionary is used to extract features from labeled data and train the classification model to realize the automatic weeds identification. In this process, this paper focuses on the effect of parameters such as the clustering number to identification accuracy under single-layer network structure, and the identification accuracy between the single-layer and the two-layer network structure was compared and analyzed. Experimental results show that identification rate can be improved by increasing the network levels, as well as fine-tuning the parameters under the premise of selecting reasonable parameters.
KeywordsFeature learning K-means clustering algorithm Feature dictionary Weeds identification
This study was funded by Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing (Grant No. 2016CP01), Xi’an Science and Technology Plan Projects (Grant No. NC1504(2)), the National Natural Science Foundation of China (Grant Nos. 31101075, 61402375), Natural Science Fundamental Research Plan of Shaanxi Province (Grant No. 2016JM6038), Fundamental Research Funds for the Central Universities, NWSUAF (Grant No. 2452015060).
Compliance with ethical standards
Conflict of interest
We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.
This article does not contain any studies with human participants or animals performed by any of the authors.
Informed consent was obtained from all individual participants included in the study.
- Agrawal KN, Singh K, Bora GC, Lin D (2012) Weed recognition using image-processing technique based on leaf parameters. J Agric Sci Technol 2(8):899–908Google Scholar
- Chong W, Blei D, Li FF (2009) Simultaneous image classification and annotation. In: Computer science department, vol 19(2), pp 1903–1910Google Scholar
- Coates A, Ng AY (2011) Selecting receptive fields in deep networks. Adv Neural Inf Process Syst 24:2528–2536Google Scholar
- Coates A, Ng AY (2012) Learning feature representations with K-means. Lecture Notes in Computer Science, vol 7700, pp 561–580Google Scholar
- Coates A, Ng AY, Lee H (2011) An analysis of single-layer networks in unsupervised feature learning. J Mach Learn Res 15:215–223Google Scholar
- D’Angelo G, Rampone S, Palmieri F (2016) Developing a trust model for pervasive computing based on Apriori association rules learning and Bayesian classification. Soft Comput 21:1–19Google Scholar
- Guan L, Liu ZZ, Wu QF, Wang LL (2016) Multi-type feature fusion technique for weed identification in cotton fields. Int J Signal Process 9(2):355–368Google Scholar
- Huang SK, Qi LJ, Zhang JH, Wang J (2013) An identification algorithm of weeds among multi-row corn based on the mapping of the corn row’s width. J China Agric Univ 18(01):165–171Google Scholar
- Kodinariya TM, Makwana PR (2013) Review on determining number of cluster in K-means clustering. Int J Adv Res Comput Sci Manag Stud 1(6):90–95Google Scholar
- Kotsiantis SB, Kanellopoulos D, Pintelas PE (2006) Data preprocessing for supervised leaning. Enformatika 1(1):111–117Google Scholar
- Krizhevsky A, Sutskever I, Hinton GE (2012) ImageNet classification with deep convolutional neural networks. Adv Neural Inf Process Syst 25(2):1097–1105Google Scholar
- Le QV (2011) Building high-level features using large scale unsupervised learning. Acoustics. In: IEEE international conference on speech and signal processing, Vancouver, pp 8595–8598Google Scholar
- Li YJ, Ao DG (2012) A fast algorithm for approximate pattern matching. J South China Univ Technol Nat Sci Edit 40(6):103–108Google Scholar
- Li XF, Zhu WX, Hua XM, Kong LD (2011) Multi-feature decision fusion method based on D-S evidence theory for apple grading. Trans Chin Soc Agric Mach 42(06):188–192Google Scholar
- Li H, Qi LJ, Zhang JH, Ji RH (2012) Recognition of weed during cotton emergence based on principal component analysis and support vector machine. Trans Chin Soc Agric 43(09):188–192Google Scholar
- Li P, He D, Qiao Y (2013) An application of soft sets in weed identification. In: American society of agricultural and biological engineers annual international meeting, vol 5, pp 4279–4288Google Scholar
- Saxe A, Bhand M, Mudur R, Suresh B, Ng AY (2011) Unsupervised learning models of primary cortical receptive fields and receptive field plasticity. Adv Neural Inf Process Syst 24:1971–1979Google Scholar
- Uetz R, Behnke S (2009) Large-scale object recognition with CUDA-accelerated hierarchical neural networks. In: IEEE international conference on intelligent computing and intelligent systems, Shanghai, pp 536–541Google Scholar
- Vincent P, Larochelle H, Bengio Y, et al (2008) Extracting and composing robust features with denoising autoencoders. In: Proceedings of the 25th international conference on machine learning, New York, pp 1096–1103Google Scholar
- Xiao CY, Zhu WX (2007) Threshold selection algorithm for image segmentation based on Otsu rule and image entropy. Comput Eng 33(14):188–189Google Scholar
- Yang Z (2014) Research on computer vision feature representation and learning for image classification and recognition. South China University of Technology, GuangzhouGoogle Scholar
- Yu K, Yang J, Gong Y (2013) Linear spatial pyramid matching using sparse coding for image classification. In: IEEE conference on computer vision and pattern recognition, Miami, pp 1794–1801Google Scholar
- Zhang K, Wang XY, Wang ZQ (2011) Hazard and control techniques of weeds in the field. Mod Agric 09:40–41Google Scholar
- Zhao B, Wang ZJ, Zhou P, Mao WH, Zhang XC (2012) Design and experiment of intelligent weed recognition system. Trans Chin Soc Agric Eng 28(s2):184–187Google Scholar
- Zhao C, He D, Qiao Y (2013) Identification method of multi-feature weed based on multi-spectral images and data mining. Trans Chin Soc Agric Eng 29(2):192–198Google Scholar
- Zhou Q, Zhao Y (2013) The design and implementation of intrusion detection system based on data mining technology. J Am Ceram Soc 92(1):204–208Google Scholar