Precisely monitoring the range of rice cultivation is an essential task for the government to dynamically supervise the red line of 180 million mu(1 mu≈666.667 m^(2))of arable land.This study aims to address the issu...Precisely monitoring the range of rice cultivation is an essential task for the government to dynamically supervise the red line of 180 million mu(1 mu≈666.667 m^(2))of arable land.This study aims to address the issues of low efficiency,high cost,and insufficient accuracy in traditional rice cultivation range monitoring methods.Against the backdrop of the widespread application of UAV remote sensing and the maturity of deep learning technology,this paper constructs a high-precision UAV remote sensing image dataset for rice identification,which includes different growth stages of rice,different resolutions,and regions.It also utilizes deep learning semantic segmentation technology to study the models,remote sensing image resolutions,and model sample sizes suitable for precise monitoring of rice.The experimental results show that,on the basis of balancing cost,efficiency,and accuracy,the Deeplabv3+and PSPNet models combined with remote sensing image data of 8 cm resolution are more suitable for monitoring and extraction of rice cultivation areas,and PSPNet has a stronger few-shot learning ability.In response to the strong model generalization ability under the dispersed rice cultivation areas and diversified features,this paper proposes a method of transfer learning with a small number of samples.This method has a more stable training process,and the IoU is 5%∼10%higher than that of unsupervised transfer learning models and fully supervised models with a small number of samples.展开更多
新型城镇化的建设和管理经常涉及到多源异构时空数据的集成和协同应用,其中基础地理数据因具有数据描述信息可被快速识别,有些数据如实时感知数据等,因没有描述信息而难以被识别,使得数据无法及时共享应用。因此,本文分析了不同新型城...新型城镇化的建设和管理经常涉及到多源异构时空数据的集成和协同应用,其中基础地理数据因具有数据描述信息可被快速识别,有些数据如实时感知数据等,因没有描述信息而难以被识别,使得数据无法及时共享应用。因此,本文分析了不同新型城镇化途径下时空数据共享应用需求,总结了时空数据的时间属性、空间属性和内容属性在共享应用中的作用,并从其可共享性快速识别的角度出发,以时空数据的时间、空间和内容等属性为主要内容设计了一个轻量级的时空数据共享信息描述框架(description framework of spatial data sharing information,DFSDSI)实现多标准时空本体数据共享特征的统一表达,从而为城镇多源异构时空数据可共享性的快速识别提供判断依据。展开更多
In aerial photography, the primary factor is terrain undulation. However, most of the external aerial photography software used for aerial photography design do not take terrain undulation influence into consideration...In aerial photography, the primary factor is terrain undulation. However, most of the external aerial photography software used for aerial photography design do not take terrain undulation influence into consideration. Therefore, the design result has comparative randomicity and ＂gaps＂ are expected. An aerial photography design system is developed by analyzing the terrain undulation influence to the design result with DEM data so that the forward overlap and side overlap can be justified according to the block terrain undulation to meet specifications or standards. The data designed by this system is compared with the real flying data. The results show that making use of DEM to assist in aerial photography design can ensure that the designed result fits the real terrain better.展开更多
基金Key Research and Development Program of Haikou(No.2022-15)Science and Technology Innovation Program of Hainan Administration of Surveying Mapping and Geo-Information,MNR(No.202406).
文摘Precisely monitoring the range of rice cultivation is an essential task for the government to dynamically supervise the red line of 180 million mu(1 mu≈666.667 m^(2))of arable land.This study aims to address the issues of low efficiency,high cost,and insufficient accuracy in traditional rice cultivation range monitoring methods.Against the backdrop of the widespread application of UAV remote sensing and the maturity of deep learning technology,this paper constructs a high-precision UAV remote sensing image dataset for rice identification,which includes different growth stages of rice,different resolutions,and regions.It also utilizes deep learning semantic segmentation technology to study the models,remote sensing image resolutions,and model sample sizes suitable for precise monitoring of rice.The experimental results show that,on the basis of balancing cost,efficiency,and accuracy,the Deeplabv3+and PSPNet models combined with remote sensing image data of 8 cm resolution are more suitable for monitoring and extraction of rice cultivation areas,and PSPNet has a stronger few-shot learning ability.In response to the strong model generalization ability under the dispersed rice cultivation areas and diversified features,this paper proposes a method of transfer learning with a small number of samples.This method has a more stable training process,and the IoU is 5%∼10%higher than that of unsupervised transfer learning models and fully supervised models with a small number of samples.
文摘新型城镇化的建设和管理经常涉及到多源异构时空数据的集成和协同应用,其中基础地理数据因具有数据描述信息可被快速识别,有些数据如实时感知数据等,因没有描述信息而难以被识别,使得数据无法及时共享应用。因此,本文分析了不同新型城镇化途径下时空数据共享应用需求,总结了时空数据的时间属性、空间属性和内容属性在共享应用中的作用,并从其可共享性快速识别的角度出发,以时空数据的时间、空间和内容等属性为主要内容设计了一个轻量级的时空数据共享信息描述框架(description framework of spatial data sharing information,DFSDSI)实现多标准时空本体数据共享特征的统一表达,从而为城镇多源异构时空数据可共享性的快速识别提供判断依据。
文摘In aerial photography, the primary factor is terrain undulation. However, most of the external aerial photography software used for aerial photography design do not take terrain undulation influence into consideration. Therefore, the design result has comparative randomicity and ＂gaps＂ are expected. An aerial photography design system is developed by analyzing the terrain undulation influence to the design result with DEM data so that the forward overlap and side overlap can be justified according to the block terrain undulation to meet specifications or standards. The data designed by this system is compared with the real flying data. The results show that making use of DEM to assist in aerial photography design can ensure that the designed result fits the real terrain better.
