


摘 ?要:為提升未知地理圖像信息標注效果,解決地理信息標注中存在的精度較低的問題,該文提出設計一種基于機器學習的地理信息協同標注方法。通過度量具體的地理信息,對已知地理信息數據進行基于機器學習的PFLP協同計算,并通過構建權矩陣,提升地理信息的流形重構與檢索標注,完成地理信息協同標注方法設計。實驗結果表明:采用基于機器學習的協同地理信息標注方法,降低了地理圖像信息特征的維數,準確展現了地理圖像數據之間的內在關系,有效地提升了地理信息協同標注的精度。
關鍵詞:機器學習 ??地理信息 ??協同標注 ?權矩陣
中圖分類號:?TP301.6 ????文獻標識碼:A???文章編號:1672-3791(2022)05(b)-0000-00
作者簡介:郝梁(1983—),男,本科,工程師,研究方向為地圖制圖。
Research on Geographic Information Collaborative Annotation Method Based on Machine learning
HAO Liang
(61243 Troops, Urumqi, Xinjiang Uygur Autonomous Region, 830006 China)
Abstract: In order to improve the effect of unknown Geographic Image Information annotation and solve the problem of low accuracy in geographic information annotation, a geographic information collaborative annotation method based on machine learning is proposed in this paper.?By measuring the specific geographic information, PFLP collaborative computing based on machine learning is carried out for the known geographic information data, and the manifold reconstruction and retrieval annotation of geographic information are improved by constructing the weight matrix, so as to complete the design of geographic information collaborative annotation method. The experimental results show that the collaborative geographic information annotation method based on machine learning reduces the dimension of geographic image information features, accurately shows the internal relationship between geographic image data, and effectively improves the accuracy of geographic information collaborative annotation.
Key Words: Machine learning; Geographic information; Collaborative annotation; Weight matrix
地理信息協同標注(GIS)是一種聚集了多種學科的新型科學。其中,包含了計算機應用、地理測繪、環境空間科學、信息與地質管理科學等技術的應用。通過電子計算機工具的應用,探索地理空間數據信息,并利用多種技術對地理位置信息進行空間操作,分析地理空間位置信息,根據確定的空間信息進行科學災害監測、地理工程研究等相關事業的開展。其中,地理信息標注過程中,由于外界多種因素的影響,導致其標注的精度較低,且工作效率較低。現有標注方法中普遍以圖形圖像、文本表格和數字等形式進行地理信息標注,這些地理信息標注方法需要耗費大量的人工效力,相對來說效率地下,對于視頻和圖像的地理信息標注檢索比較困難。……