|
|
|
|
|
|
Overlapping Green Apple Recognition Based on Improved Spectral Clustering |
LI Da-hua1, ZHAO Hui2*, YU Xiao3 |
1. School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, China
2. School of Engineering and Technology, Tianjin Agricultural University, Tianjin 300384, China
3. School of Electrical and Electronic Engineering, Tianjin University of Technology, Tianjin 300384, China |
|
|
Abstract Fruits target recognition is one of the most important steps to realize agricultural automation. In the process of fruit recognition, because of the influence of overlap and occlusion, the target recognition is difficult, and the rate of recognition is not high. The paper uses spectral clustering algorithm to solve the problem of overlapped fruit in natural environment. Then the identification and location of fruit are realized by randomized hough transform. In view of the large number of computation and the slow operation speed of the traditional algorithm, this paper proposes an improved spectral clustering algorithm based on Mean Shift and sparse matrix principle. Firstly, the image is pre-segmented using the mean shift algorithm. Mean shift is a non-parametric estimation method for density gradient. The algorithm is essentially an iteration. Calculate the offset, move the point according to the offset, and repeat the above steps until the offset is zero. Most of the background pixels are removed by mean shift algorithm, and the removing is prepared for reducing the computational complexity of the spectral clustering algorithm. And then the useful information is extracted, which is the description of the similarity between the pairs of pixels in the image, and the extracted image feature information is mapped into a sparse matrix. The K-means algorithm is used to classify it into classes, and the final classification result is obtained to realize the re-segmentation of the reprocessed image. Then the color of the image segmentation area is restored, the edge contour is extracted by using a color vector gradient and the randomized hough transform is used on the resulting contour image, and the radius parameter range during the detection process is set to further accelerate the speed of the algorithm. The center coordinates and radius of the target can be obtained through the detection. Thereby the overlapped green apples are recognized. Finally, the algorithm has the high coincidence degree of 95.41%, the low error rate of 4.59% and the false detection rate of 3.05% through experimental analysis and algorithm comparison, and the algorithm meets the practical application requirements.
|
Received: 2018-07-27
Accepted: 2019-01-05
|
|
Corresponding Authors:
ZHAO Hui
E-mail: zhaohui3379@126.com
|
|
[1] Niu L, Zhou W, Wang D, et al. Multimedia Tools & Applications,2016. 1.
[2] Song Huaibo, Zhang Chuantong, Pan Jingpeng, et al. Journal of Agricultural Engineering,2013,(3): 163.
[3] Xu Yue, Li Yinghui, Song Huaibo, et al. Transactions of the Chinese Society of Agricultural Engineering,2015, 31(1): 196.
[4] Arefi A, Motlagh A M, Mollazade K, et al. Australian Journal of Crop Science, 2011, 5(10): 1144.
[5] Cong Peisheng, Sun Jianzhong,Chinese Journal of Image Graphics, 2006, 11(12): 1781.
[6] Xie Zhonghong, Ji Changying, Guo Xiaoqing, et al. Journal of Agricultural Machinery,2011, 42(12): 191.
[7] Rong Yingbin, Jiang Huanyu, et al. Journal of Agricultural Machinery,2012, 43(3): 157.
[8] Wang Dandan, Xu Yue, Song Huaibo, et al. Transactions of the Chinese Society of Agricultural Engineering,2015, (10): 227.
[9] Petrus B, Haakan L, Eric A. IEEE Transactions on Information Theory,1975, 21(1): 32.
[10] Tao W, Jin H, Zhang Y. IEEE Transactions on Systems Man & Cybernetics Part B,2007, 37(5): 1382.
[11] Huo Hua, Feng Boqin. Minicomputer System,2005, 26(6): 988.
[12] Li Dahua, Shen Mingming, Li Dong, et al. IEEE International Conference on Mechatronics and Automation, 2017. 264.
[13] Ai Lingyu, Kim Eun-Soo. Optics Communications, 2018, 410: 871. |
[1] |
LI Xin-li1, CONG Li-li2, XU Shu-ping2, LI Su-yi1*. Cell Growth Analysis Method Based on Spectral Clustering and Single-Cell Raman Spectroscopy[J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2023, 43(09): 2832-2836. |
[2] |
ZHOU Meng-ran, SONG Hong-ping*, HU Feng, LAI Wen-hao, WANG Jin-guo. Application of Spectral Clustering and LIF in Recognition of Mine Water Inrush Source Types[J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2021, 41(02): 435-440. |
[3] |
QIN Fang-pu1, 2 , ZHANG Ai-wu1, 2*, WANG Shu-min3, MENG Xian-gang1, 2, HU Shao-xing4, SUN Wei-dong5 . Hyperspectral Band Selection Based on Spectral Clustering and Inter-Class Separability Factor [J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2015, 35(05): 1357-1364. |
[4] |
YU Bo1, 2, NIU Zheng1*, WANG Li1, LIU Ya-qi3, CHEN Fang4 . An Unsupervised Method of Extracting Constructions from Color Remote Sensed Image Based on Mean Shift and Neutrosophic Set [J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2013, 33(04): 1071-1075. |
[5] |
WANG Lei-guang1,3, ZHENG Chen2, LIN Li-yu3*, CHEN Rong-yuan3, MEI Tian-can4 . Fast Segmentation Algorithm of High Resolution Remote Sensing Image Based on Multiscale Mean Shift [J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2011, 31(01): 177-183. |
[6] |
CHEN Yu1, 2, LIN Qi-zhong1, WEI Yong-ming3, WANG Meng-fei3, LI Hui1, 2. Alteration Mineral Mapping Method Based on Statistical Analysis of Field Measured Spectra[J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2010, 30(11): 3036-3040. |
|
|
|
|