印度南瓜单瓜产籽数与影响因素分析

通过对印度南瓜果实主要数量性状的通径分析表明，各性状与单瓜产籽数的相关程度依次为单瓜重（０．８７６５）〉瓜横径（０．８６２２）〉瓜纵径（－０．７４７２）〉籽粒长（０．５６９４）〉籽粒宽（０．４２８４）〉千粒重（－０．０６０４）。各性状对单瓜产籽数起直接作用大小依次为瓜横径（０．４７４９）〉单瓜重（０．３６３８）〉籽粒宽（０．２１２２）〉瓜纵径（－０．１７４１）〉籽粒长（－０．１２５７）〉千粒重（－     

改进C-V模型在YCbCr空间下的羊体图像分割

针对羊体图像背景复杂、分割难以及不同光照条件干扰羊体图像的问题,采用一种基于YCbCr空间改进C-V主动轮廓模型的分割方法,对具有复杂背景的羊体图像分割进行研究。结果表明:1)根据羊体图像的颜色特点,对羊体图像进行从RGB空间到YCbCr空间的转换能克服拍摄环境中光照对羊体的影响;2)利用手动勾画羊体的粗略轮廓构造预处理水平集,对其内部、外部以及边界进行划分后可以演化羊体图像的轮廓。试验证明改进C-V模型能对复杂背景下的羊体图像进行准确分割,分割结果能够应用到后续羊体测量点的识别中。     

基于改进区域生长法的羊体点云分割及体尺参数测量

针对传统羊体尺测量中测量耗时、应激大的问题,采用主成分分析、随机采样一致性算法和改进的区域生长法,基于三维点云对羊体尺参数测量进行研究。结果表明:1)使用主成分分析和随机采样一致性算法能计算羊体点云的法向量和曲率;2)改进的区域生长法能准确地分割出羊体区域,并且避免了外点的干扰;3)在羊体点云数据上选取体尺测点,计算羊体长、体高、臀高、胸深体尺参数,并与实测值比较,4种体尺参数的最大相对误差为2.36%,测量精度较高。试验证明改进的区域生长法能准确地对羊体点云进行分割,依据选取的体尺测点,能够实现羊体尺参数的无接触测量。     

Mapping and identifying basal stem rot disease in oil palms in North Sumatra with QuickBird imagery

The application of remote sensing technology and precision agriculture in the oil palm industry is in development. This study investigated the potential of high resolution QuickBird satellite imagery, which has a synoptic overview, for detecting oil palms infected by basal stem rot disease and for mapping the disease. Basal stem rot disease poses a major threat to the oil palm industry, especially in Indonesia. It is caused by Ganoderma boninense and the symptoms can be seen on the leaf and basal stem. At present there is no effective control for this disease and early detection of the infection is essential. A detailed, accurate and rapid method of monitoring the disease is needed urgently. This study used QuickBird imagery to detect the disease and its spatial pattern. Initially, oil palm and non oil palm object segmentation based on the red band was used to map the spatial pattern of the disease. Secondly, six vegetation indices derived from visible and near infrared bands (NIR) were used for to identify palms infected by the disease. Finally, ground truth from field sampling in four fields with different ages of plant and degrees of infection was used to assess the accuracy of the remote sensing approach. The results show that image segmentation effectively delineated areas infected by the disease with a mapping accuracy of 84%. The resulting maps showed two patterns of the disease; a sporadic pattern in fields with older palms and a dendritic pattern in younger palms with medium to low infection. Ground truth data showed that oil palms infected by basal stem rot had a higher reflectance in the visible bands and a lower reflectance in the near infrared band. Different vegetation indices performed differently in each field. The atmospheric resistant vegetation index and green blue normalized difference vegetation index identified the disease with an accuracy of 67% in a field with 21 year old palms and high infection rates. In the field of 10 year old palms with medium rates of infection, the simple ratio (NIR/red) was effective with an accuracy of 62% for identifying the disease. The green blue normalized difference vegetation index was effective in the field of 10 years old palms with low infection rates with an accuracy of 59%. In the field of 15 and 18 years old palms with low infection rates, all the indices showed low levels of accuracy for identifying the disease. This study suggests that high resolution QuickBird imagery offers a quick, detailed and accurate way of estimating the location and extent of basal stem rot disease infections in oil palm plantations.     

基于特征匹配的羊体点云配准方法

针对点云配准算法易受噪声、体外孤点以及采样率影响的问题,采用形状指数关键点检测方法、最近邻距离比法和迭代最近点算法,基于三维点云对羊体点云配准方法进行研究。结果表明:1)使用协方差矩阵特征描述子能对形状指数方法检测的关键点进行描述;2)基于特征匹配的配准方法能对不同视角的羊体点云进行配准,最大均方根误差为0.024 1;3)对于含有噪声、体外孤点或较低采样率的不同类型的羊体点云模型,配准的最大均方根误差为0.023 8。试验证明基于特征匹配的配准方法能准确地对羊体点云进行配准,并且不受噪声、体外孤点以及采样率的影响。     

Impact of intensive agricultural management on carbon and nitrogen dynamics in the humid tropics

ABSTRACT

The concern about global climate change continues to increase research interest regarding carbon and nitrogen dynamics in the soil. This is based on their role in maintaining soil fertility, which can instead be a source of greenhouse gas emissions if not managed properly, while threatening food security. Humid tropical conditions enable intensive agricultural cultivation with various cropping systems to fulfill the demand for agriculture products. Such climate accelerates the soil organic matter decomposition rate so that it strongly influences soil carbon and nitrogen dynamics. However, inappropriate implementation of intensive agricultural systems that does not consider the balance between carbon and nitrogen input and output, negatively affects soil fertility, mainly decreasing soil organic carbon and total soil nitrogen, changing the composition of carbon and nitrogen owing to the loss of soil organic matter through erosion and leaching, thus, causing soil degradation. Mitigation strategies can be performed by using organic matter and crop residue, crop rotation and improvement of crop pattern, soil tillage and fertilization, cover crops and mulch. Sustainable land management for maintenance of soil organic carbon and total soil nitrogen dynamics should be locally and globally developed and adopted for a more sustainable agricultural system. Recovery of soil capacity to accumulate carbon is a strategic step to reduce the impact of climate change. Hence, an intensive study on efficient soil organic carbon management is required to improve food production and mitigation of climate change to attain sustainable development goals in 2030.     

基于多尺度Retinex图像增强的羊体尺参数无接触测量

针对传统羊体尺参数测量中存在工作量大、精度低、应激反应强等问题,采用多尺度Retinex算法、Graph Cut算法和羊体尺测点识别相结合的方法,基于双目视觉检测原理对羊体尺参数的无接触测量进行研究。结果表明:1)带色彩恢复的多尺度Retinex算法能增强光照不均匀的羊图像,对羊图像的细节和颜色恢复表现出较强的处理能力;2)基于多尺度分水岭的Graph Cut算法准确地分割出羊体区域,获得光滑的羊体轮廓线;3)划分羊体轮廓线区域,采用包络线分析方法识别体尺测点,计算羊体尺参数,并与真实值比较,11只羊的平均相对误差为2.32%,除去绒山羊剩下9只羊的平均相对误差为1.95%,测量精度较高。试验证明本研究带色彩恢复的多尺度Retinex算法能增强羊图像亮度和色度,包络线分析方法能准确地识别体尺测点,算法稳定,能够实现饲养过程中羊体尺参数的无接触测量。     

Smallholder Farmers’ Knowledge of Regulations Governing the Sale of Timber and Supply Chains in Gunungkidul District,Indonesia

Smallholder timber plantations may offer opportunities for farmers to increase their income. Nonetheless, such opportunities are often lost largely due to unfavorable regulations imposed on harvesting and marketing of timber. Adverse impacts are worsened because the regulations are not effectively communicated to smallholder farmers. We assessed the level of smallholder knowledge of existing regulations and found it very low. In part, this correlated to their socio-demographic characteristics. Poor literacy skills and a low level of education make it difficult for older farmers’ to increase their knowledge. This is compounded by the limited information channels that can reach the farmers. Improving smallholders’ knowledge by providing information concerning markets and regulations in a timely and clear manner could help smallholders exercise coping strategies and priorities when selling their timber that would in turn reduce the negative impacts of regulations. The roles of village authorities could be crucial. They are close to the smallholders and could channel the information. The government could provide training and incentives for them to be more active in communicating the regulations to the smallholders.