Abstract: In order to understand the evolution of hyperspectral technology in the field of agricultural quantitative modeling research at home and abroad, and to contribute to the practicality of precision agriculture, a dataset comprising 805 Chinese and 1 115 English academic papers on agricultural hyperspectral quantitative modeling from the CNKI and WOS databases spanning the years of 2007—2022 was utilized. Combining this data with the visualization and bibliometric analysis software CiteSpace, the study examined trends in publication volume, contributing journals, research institutions, authors, and collaborative relationships over the 15‑year period. The analysis revealed a gradual increase in the number of publications in WOS, surpassing those in CNKI. The top journals, both in Chinese and English, accounted for 12% and 19% of the total literature, respectively. Among the top five international research institutions, the United States Department of Agriculture had the highest publication volume, while the remaining four were Chinese research institutions. Analysis of core author groups and keywords reflected that the mainstream research direction focused on the retrieval of nutritional elements in food crops such as wheat and rice. Keyword burst analysis further highlighted research hotspots and evolutionary trends at different stages in the field. The results show that quantitative agricultural hyperspectral modeling was widely used in crop growth status detection, crop quality detection and crop yield estimation. However, there were many problems such as fusion and mapping of crop hyperspectral data at different scales, intelligence of data analysis and modeling, and adaptability of quantitative models. Combining the characteristics of different crops, integrating the advantages of different technologies, and improving the accuracy of data and models will become the key research direction in the field of agricultural hyperspectral quantitative modeling in the future.

Key words: crops, bibliometrics, hyperspectral, quantitative model

摘要： 为了解国内外高光谱技术在农业定量建模研究领域的演进历程，助力提高精准农业的实用性，以2007—2022年CNKI和WOS数据库中农业高光谱定量建模领域的中文805条，英文1 115条学术性文献为数据源，结合可视化文献计量分析软件CiteSpace统计分析15年间的发文量趋势、发文期刊及科研机构、发文作者及合作关系等，其中WOS发文量呈缓慢增长趋势，且高于CNKI；中英文刊文量最多的期刊分别占总文献数的12%、19%；国际发文量前5的科研机构中，美国农业部发文量最多，其余四席均为中国科研机构；核心作者群及关键词分析反映出主流研究方向是对小麦、水稻等粮食作物的营养元素的反演研究；通过关键词突现分析，明确该领域不同阶段的研究热点和演化趋势。结果表明，目前农业高光谱定量建模在作物生长状况检测、作物品质检测以及作物产量估测方面的应用较为广泛，但目前存在不同尺度的作物高光谱数据融合与映射、数据分析与建模的智能化以及定量模型的适应性等问题，结合不同作物特点进行数据融合、集合不同技术的优势，提高数据及模型准确性将成为未来农业高光谱定量建模研究领域的重点研究方向。

关键词: 农作物, 文献计量学, 高光谱, 定量模型