|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | 基于NB-IoT的温室温度智能调控系统设计与实现显示文摘【目的】以NB-IoT低速率窄带宽物联网技术为核心,研制一套以5G低功耗海量连接场景前期技术为基础的智能温室环境自动调控系统。【方法】应用MSP430F149超低功耗芯片采集环境信息,依托NB-IoT蜂窝物联网平台,云端智能调控系统,结合多传感器融合与模糊PID–分级控制技术,根据用户需求调节温室环境。【结果】该系统在温室大棚内实地应用的结果表明:温室环境信息采集相对误差不超1%,平均控制精度在3.57%(±1.0℃),无传输距离限制,实现作物生长温度的自动调节。【结论】该系统稳定可靠,为作物的生长提供良好环境,对作物的研究提供有力的技术支撑。 | 何灿隆 沈明霞 刘龙申 OKINDA Cedric 杨稷 施宏 | 2018 | 华南农业大学学报2018，39，2： | 49 |
| 2 | 基于深层卷积神经网络的初生仔猪目标实时检测方法显示文摘针对初生仔猪目标较小、分娩栏内光线变化复杂、仔猪粘连和硬性遮挡现象较为严重等问题,提出一种基于深层卷积神经网络的初生仔猪目标识别方法。将分类和定位合并为一个任务,以整幅图像为兴趣域,利用特征金字塔网络(Feature pyramid network,FPN)算法定位识别仔猪目标;对比了不同通道数数据集以及不同迭代次数对模型效果的影响;该方法支持图像批量处理、视频与监控录像的实时检测和检测结果多样化储存。实验结果表明:在数据集总量相同时,同时包含夜间单通道和白天3通道的数据集,在迭代20 000次时接近模型最优值。模型在验证集和测试集上的精确率分别为95.76%和93.84%,召回率分别为95.47%和94.88%,对分辨率为500像素×375像素的图像检测速度为53.19 f/s,对清晰度为720 P的视频检测速度为22 f/s,可满足实时检测的要求,对全天候多干扰场景表现出良好的泛化能力。 | 沈明霞 太猛 CEDRIC Okinda 刘龙申 李嘉位 孙玉文 | 2019 | 农业机械学报2019，50，8： | 24 |
| 3 | A computer vision system for defect discrimination and grading in tomatoes using machine learning and image processing显示文摘With large-scale production and the need for high-quality tomatoes to meet consumer and market standards criteria,have led to the need for an inline,accurate,reliable grading system during the post-harvest process.This study introduced a tomato grading machine vision system based on RGB images.The proposed system performed calyx and stalk scar detection at an average accuracy of 0.9515 for both defected and healthy tomatoes by histogramthresholding based on themean g-r value of these regions of interest.Defected regionswere detected by an RBF-SVMclassifier using the LAB color-space pixel values.Themodel achieved an overall accuracy of 0.989 upon validation.Four grading categories recognitionmodelswere developed based on color and texture features.The RBF-SVMoutperformed all the explored modelswith the highest accuracy of 0.9709 for healthy and defected category.However,the grading accuracy decreased as the number of grading categories increased.A combination of color and texture features achieved the highest accuracy in all the grading categories in image features evaluation.This proposed system can be used as an inline tomato sorting tool to ensure that quality standards are adhered to and maintained. | David Ireri Eisa Belal Cedric Okinda Nelson Makange Changying Ji | 2019 | Artificial Intelligence in Agriculture2019，，2： | 10 |
| 4 | Optimization of compression formulation and load of food-grade tracers for grain traceability using central composite design显示文摘Food-grade tracers have been developed as an identification technology for grain traceability from original harvest to final destination for transportation.The characteristics of food-grade tracers must be able to satisfy the environmental demands for grain traceability.To optimize the food-grade tracer production process,the effects of direct compression formulation and load on the mechanical characteristics were studied using response surface methodology(RSM)with central composite design(CCD).Among the four tested formulations,Formulations#2(consisting of 35.00%lactose 100 mesh,64.50%microcrystalline cellulose 102 and 0.50%magnesium stearate)and#4(consisting of 38.00%lactose 100 mesh,50.00%microcrystalline cellulose 102,11.00%pregelatinized starch and 1.00%magnesium stearate)were selected for tracer production based on their physical properties as powders.The value of Carr’s flowability index was 68 for both Formulations#2 and#4,which was the highest among all the formulations.Therefore,Formulations#2 and#4 also had the best powder flowability.The magnesium stearate ratio(1.00%-3.00%)and pressure(6.00-16.00 kgf)were used as independent variables to detect changes in the breaking rate,peak shear force and friction coefficient of tracers compressed by the selected formulations.The optimal production parameters could be achieved at a magnesium stearate ratio of 2.25%and pressure of 16.00 kgf for Formulation#2 and at a magnesium stearate ratio of 1.02%and pressure of 16.00 kgf for Formulation#4.Under these optimal conditions,the tracers had good impact characteristics(breaking rate),compression characteristics(peak shear force)and frictional characteristics(friction coefficient).Moreover,Formulation#2 was more suitable for production because compared to Formulation#4,its breaking rate and friction coefficient values were lower,and its peak shear force value was higher. | Liang Kun Zhang Lingling Lu Wei Cedric Sean Okinda Shen Mingxia | 2017 | International Journal of Agricultural and Biological Engineering2017，10，6： | 2 |
| 5 | A review on computer vision systems in monitoring of poultry: A welfare perspective显示文摘Monitoring of poultry welfare-related bio-processes and bio-responses is vital in welfare assessment and management of welfare-related factors.With the current development in information technologies,computer vision has become a promising tool in the real-time automation of poultry monitoring systems due to its non-intrusive and non-invasive properties,and its ability to present a wide range of information.Hence,it can be applied to monitor several bio-processes and bio-responses.This review summarizes the current advances in poultrymonitoring techniques based on computer vision systems,i.e.,conventional machine learning-based and deep learning-based systems.A detailed presentation on the machine learning-based system was presented,i.e.,pre-processing,segmentation,feature extraction,feature selection,and dimension reduction,and modeling.Similarly,deep learning approaches in poultry monitoring were also presented.Lastly,the challenges and possible solutions presented by researches in poultry monitoring,such as variable illumination conditions,occlusion problems,and lack of augmented and labeled poultry datasets,were discussed. | Cedric Okinda Innocent Nyalala Tchalla Korohou Celestine Okinda Jintao Wang Tracy Achieng Patrick Wamalwa Tai Mang Mingxia Shen | 2020 | Artificial Intelligence in Agriculture2020，，1： | 2 |
| 6 | Short-term feeding behaviour sound classification method for sheep using LSTM networks显示文摘A deep learning approach using long-short term memory(LSTM)networks was implemented in this study to classify the sound of short-term feeding behaviour of sheep,including biting,chewing,bolus regurgitation,and rumination chewing.The original acoustic signal was split into sound episodes using an endpoint detection method,where the thresholds of short-term energy and average zero-crossing rate were utilized.A discrete wavelet transform(DWT),Mel-frequency cepstral,and principal-component analysis(PCA)were integrated to extract the dimensionally reduced DWT based Mel-frequency cepstral coefficients(denoted by PW_MFCC)for each sound episode.Then,LSTM networks were employed to train classifiers for sound episode category classification.The performances of the LSTM classifiers with original Mel-frequency cepstral coefficients(MFCC),DWT based MFCC(denoted by W_MFCC),and PW_MFCC as the input feature coefficients were compared.Comparison results demonstrated that the introduction of DWT improved the classifier performance effectively,and PCA reduced the computational overhead without degrading classifier performance.The overall accuracy and comprehensive F1-score of the PW_MFCC based LSTM classifier were 94.97%and 97.41%,respectively.The classifier established in this study provided a foundation for an automatic identification system for sick sheep with abnormal feeding and rumination behaviour pattern. | Guanghui Duan Shengfu Zhang Mingzhou Lu Cedric Okinda Mingxia Shen Tomas Norton | 2021 | International Journal of Agricultural and Biological Engineering2021，14，2： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员