长绒棉棉铃成熟度仿生类脑分类方法研究

doi:10.13733/j.jcam.issn.2095-5553.2024.10.026

摘要/Abstract

摘要： 快速准确识别长绒棉棉铃成熟的不同阶段，对长绒棉种植的智能化、装备化管理具有重要意义。针对现有方法在长绒棉棉铃成熟度分类辨识易受复杂棉田背景、阴影、强光和叶片遮挡等因素影响识别率低的问题，提出将模拟生物视觉皮层的信息处理机制与HMAX模型融合的类脑分类方法，实现大田环境下棉铃不同成熟度的快速、高效识别。首先，采用相机获取长绒棉不同成熟阶段的图像信息，以花铃期、裂铃期、吐絮期、停止生长期4个生长阶段为对象，构建中小样本的棉铃数据集；其次，模拟视网膜神经节细胞的信息处理机制以提升HMAX模型的检测速度与精度，提出基于改进HMAX模型的类脑识别算法；最后，为探究各模型在非清晰数据集上的表现，采用高斯模糊方法将测试集转换6次，以HMAX、HHMAX、SHMAX作为对比，评估改进HMAX模型性能。试验结果表明，在原始数据集下，改进HMAX模型的总体准确率为95.3%，相比于HMAX、HHMAX、SHMAX模型分别高出15.1、9.2和6个百分点；在错误分类中，由于吐絮期和停止生长期特征相似，造成错误识别的概率最大；在非清晰数据集下，HMAX、HHMAX以及SHMAX退化指数分别为8.21%、7.935%和11.21%，改进HMAX模型总体退化指数为5.92%。研究结果表明：在分类精度、模糊图像输入等方面，改进HMAX模型能够较好地满足实际生产中棉铃不同成熟阶段分类识别的实际需求。

关键词: 棉铃成熟度, 分类识别, 改进HMAX模型, 图像处理, 仿生类脑

Abstract: Rapid and accurate identification of different stages of long‑staple cotton boll maturity is important for intelligent and equipped management of long‑staple cotton cultivation. In order to address the problem that the existing methods in long‑staple cotton boll maturity classification recognition are easily affected by complex cotton field backgrounds, shadow, bright light and leaf shading, this paper proposes a brain‑like classification method that integrates the information processing mechanism of simulated biological visual cortex with the HMAX model to achieve fast and efficient recognition of different boll maturity in a large field environment. Firstly, a camera is used to obtain the image information of different maturity stages of long‑staple cotton, and four growth stages, namely, the boll stage, boll splitting stage, linting stage and cessation stage, are used to build a small and medium‑sized boll dataset. Secondly, the information processing mechanism of retinal ganglion cells is simulated to improve the detection speed and accuracy of the HMAX model, and a brain‑like recognition algorithm based on the improved HMAX model is proposed. Finally, in order to explore the performance of each model on non‑clear data sets, the test set was transformed six times by using the Gaussian fuzzy method, and HMAX, HHMAX and SHMAX were used as comparisons to evaluate the performance of the improved HMAX model. The experimental results showed that the overall accuracy of the improved HMAX model on the original test set was 95.3%, which was 15.1, 9.2 and 6 percentage points higher, respectively, compared with the HMAX, HHMAX and SHMAX models. In the misclassification, the probability of false identification was the highest due to the similarity of the characteristics of the situation and cessation growth periods. Under the non‑clear data set, the degradation indices of HMAX, HHMAX and SHMAX were 8.21%, 7.935%, and 11.21%, respectively, and the overall degradation index of the improved HMAX model was 5.92%. The results show that the improved HMAX model can better meet the practical needs of classification and recognition of cotton boll at different maturity stages in actual production in terms of classification accuracy and fuzzy image input.

Key words: cotton boll maturity, classification and recognition, improved HMAX model, image processing, bionic brain

中图分类号:

S792.91

崔高建, 李晓娟. 长绒棉棉铃成熟度仿生类脑分类方法研究[J]. 中国农机化学报, 2024, 45(10): 177-183.

Cui Gaojian, Li Xiaojuan. Research on bionic brain‑inspired classification method for maturity of long‑staple cotton bolls#br#[J]. Journal of Chinese Agricultural Mechanization, 2024, 45(10): 177-183.

参考文献

[ 1 ] 郑子漂, 徐海江, 林涛, 等. 新疆长绒棉育成品种演变趋势及综合评价[J]. 中国农业大学学报, 2022, 27(6): 55-70.
Zheng Zipiao, Xu Haijiang, Lin Tao, et al. Evolution trend and comprehensive evaluation of growing varieties of long staple cotton in Xinjiang [J]. Journal of China Agricultural University, 2022, 27(6): 55-70.
[ 2 ] 李亚楠. 基于视觉的棉花发育状态自动检测技术研究[D]. 武汉: 华中科技大学, 2018.
[ 3 ] 段宇飞, 孙记委, 王焱清, 等. 基于改进卷积神经网络的油茶果壳籽分选方法[J]. 农业工程学报, 2023, 39(3): 154-161.
Duan Yufei, Sun Jiwei, Wang Yanqing, et al. Sorting camellia oleifera husk and seed using an improved convolutional neural network [J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(3): 154-161.
[ 4 ] 刘平, 刘立鹏, 王春颖, 等. 基于机器视觉的田间小麦开花期判定方法[J]. 农业机械学报, 2022, 53(3): 251-258.
Liu Ping, Liu Lipeng, Wang Chunying, et al. Determination method of field wheat flowering period baesd on machine vision [J]. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(3): 251-258.
[ 5 ] Zhang Y, Yang G, Liu Y, et al. An improved YOLO network for unopened cotton boll detection in the field [J]. Journal of Intelligent & Fuzzy Systems, 2022, 42(3): 2193-2206.
[ 6 ] Liu J, Lai H, Jia Z. Image segmentation of cotton based on YCbCcr color space and fisher discrimination analysis [J]. Acta Agronomica Sinica, 2011, 420(2): 274-281.
[ 7 ] 徐建鹏, 王杰, 徐祥, 等. 基于RAdam卷积神经网络的水稻生育期图像识别[J]. 农业工程学报, 2021, 37(8): 143-150.
Xu Jianpeng, Wang Jie, Xu Xiang, et al. Image recognition for different developmental stages of rice by RAdam deep convolutional neural networks [J]. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(8): 143-150.
[ 8 ] Chen Q Z, Lai H C, Wang X, et al. An cotton image segmentation algorithm based on support vector machine [J]. Computer Engineering, 2013, 39(5): 266-269.
[ 9 ] Li Y, Cao Z, Lu H, et al. In‑field cotton detection via region‑based semantic image segmentation [J]. Computers and Electronics in Agriculture, 2016, 127: 475-486.
[10] Sun S, Li C, Paterson A H, et al. Image processing algorithms for infield single cotton boll counting and yield prediction [J]. Computers and Electronics in Agriculture, 2019, 166: 104976.
[11] Patrício D I, Rieder R. Computer vision and artificial intelligence in precision agriculture for grain crops: A systematic review [J]. Computers and Electronics in Agriculture, 2018, 153: 69-81.
[12] Xu R, Li C, Paterson A H, et al. Aerial images and convolutional neural network for cotton bloom detection [J]. Frontiers in Plant Science, 2018, 8: 2235.
[13] Riesenhuber M, Poggio T. Hierarchical models of object recognition in cortex [J]. Nature Neuroscience, 1999, 2(11): 1019-1025.
[14] Serre T, Wolf L, Bileschi S, et al. Robust object recognition with cortex‑like mechanisms [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(3): 411-426.
[15] Shariatmadar Z S, Faez K. A data‑driven and biologically inspired preprocessing scheme to improve visual object recognition [J]. Computational Intelligence and Neuroscience, 2021, 2021: 1-13.
[16] Ruigang F, Biao L, Yinghui G, et al. Accelerated HMAX model based on improved SIFT feature points [C]. 2015 IEEE International Conference on Grey Systems and Intelligent Services (GSIS). IEEE, 2015: 485-489.
[17] Norizadeh Cherloo M, Shiri M, Daliri M R. An enhanced HMAX model in combination with SIFT algorithm for object recognition [J]. Signal, Image and Video Processing, 2020, 14: 425-433.
[18] Sufikarimi H, Mohammadi K. Role of the secondary visual cortex in HMAX model for object recognition [J]. Cognitive Systems Research, 2020, 64: 15-28.

[1]	杨浩勇 , 王超柱 , 刘浩义 , 关心桐 , 刘璎瑛 , 丁永前 . 种子破损率快速检测方法研究[J]. 中国农机化学报, 2024, 45(9): 104-110.
[2]	张倩, 王明, 于峰, 陶震宇, 张辉, 李刚. 基于CNN的作物分类识别图像获取平台研究进展[J]. 中国农机化学报, 2024, 45(8): 170-179.
[3]	张延军, 赵建鑫, . 基于多路神经网络的黄花菜关节点检测算法研究[J]. 中国农机化学报, 2024, 45(7): 228-234.
[4]	郎秀丹, 史宇亮, 李天华, 陈明东. 基于阈值识别的温室黄瓜整体自动落蔓装置研究[J]. 中国农机化学报, 2024, 45(6): 106-112.
[5]	杨昊霖, 王其欢, 李华彪, 耿端阳, 武继达, 姚艳春. 基于改进YOLOv5的田间复杂环境障碍物检测[J]. 中国农机化学报, 2024, 45(6): 216-222.
[6]	林燕翔, 沈印, 李光林. 基于改进InceptionV3算法的小麦杂质识别研究[J]. 中国农机化学报, 2024, 45(4): 108-116.
[7]	缪亚伦, 石美琦, 孟海涛, 梁旭升, 黄才贵, 李岩舟. 基于Stereo CameraYOLOv5自然环境下百香果检测与定位模型[J]. 中国农机化学报, 2024, 45(3): 233-241.
[8]	张伟进, 王福顺, , 孙小华, 王军皓, 刘宏权, 王鑫鑫, . 传统图像分割算法在农作物籽粒考种应用中的研究进展[J]. 中国农机化学报, 2024, 45(2): 280-287.
[9]	李家超, 罗斌, 周亚男, 黄硕, 侯佩臣, . 基于图像处理的水稻剑叶夹角测量[J]. 中国农机化学报, 2024, 45(10): 228-232.
[10]	方俊泽, 郭正, 李歌, 邢素霞, 王瑜. 基于改进Swin-Transformer的柑橘病叶分类模型[J]. 中国农机化学报, 2024, 45(1): 252-258.
[11]	黄家才, 汪涛, 张铎, 唐安, 高芳征. 基于改进YOLOv4的草莓果柄叶识别与定位[J]. 中国农机化学报, 2023, 44(9): 205-213.
[12]	李航, 廖映华, 黄波. 基于改进DQN算法的茶叶采摘机械手路径规划[J]. 中国农机化学报, 2023, 44(8): 198-205.
[13]	周显沁, 韩震宇, 刘成源. 基于改进卷积神经网络与图像处理的蚕茧识别方法[J]. 中国农机化学报, 2023, 44(5): 100-106.
[14]	周鹏飞, 蒙贺伟, , , 梁荣庆, 张炳成, 坎杂, , . 基于OpenCV-Python的土壤颗粒动态休止角试验[J]. 中国农机化学报, 2023, 44(5): 214-222.
[15]	王恒, 张维懿, 李成松, 王沛, 王丽红, . 基于图像处理的果树滴水线导航路径检测方法研究[J]. 中国农机化学报, 2023, 44(3): 183-190.