Abstract: ‍Single object tracking technology is a fundamental research problem in the field of computer vision， with extensive applications across various domains. Single object tracking methods based on RGB images achieved remarkable success following years of meticulous development. However， these methods exhibit substantial limitations and challenges when deployed in low-light environments， cluttered backgrounds， and scenarios with interference from similar targets. By contrast， infrared thermal imaging and hyperspectral imaging enhance object tracking technologies by providing thermal radiation characteristics and spectral features， thereby extending the spectral information of target representation. Despite the great progress in single object tracking methods with extended spectral information based on RGB-T（RGB-Thermal） and hyperspectral images， a comprehensive review of the literature remains lacking. To bridge this gap， this paper aims to provide a thorough review of single object tracking technologies with extended spectral information. First， this paper presents a survey of the domestic and international development of single object tracking technologies enhanced by extended spectral features， reviews the frameworks based on RGB images， and summarizes those based on RGB-T and hyperspectral images. The frameworks are categorized into those based on traditional， Siamese-based， and Transformer-based tracking methods. With the advancements of computational capabilities and the development of deep learning， deep learning-based tracking methods are increasingly demonstrating exceptional performance. Further， this paper provides a detailed discussion on the methods for improving object tracking based on RGB-T images， including the design of feature fusion strategies， utilization of prior challenge attributes in video sequences， and the generation of high-quality feature representation. Moreover， this paper extensively describes the structure， improvement objectives， and methodological advantages of object tracking methods based on hyperspectral images， covering aspects such as traditional hyperspectral tracking methods， the construction of spectral band selection strategies， efficient expression of hyperspectral features， and the design of universal models suitable for various hyperspectral imaging devices. Furthermore， the current datasets and evaluation metrics are compiled based on RGB-T and hyperspectral imaging， the performance of various tracking algorithms is compared for these datasets， and the advantages and disadvantages of the different methods are summarized. Among these， novel RGB-T tracking methods based on generative models achieve impressive tracking results. Finally， this study explores the future trends of single object tracking methods with extended spectral information. The conclusions are summarized as follows. Single object tracking methods based on RGB-T and hyperspectral images attract significant interest among researchers and are poised to be crucial in the future of object tracking technologies. The available RGB-T and hyperspectral tracking datasets are relatively limited in volume and diversity， which restricts the training and evaluation of algorithms； hence， constructing large-scale datasets is fundamental for advancing tracking methodologies. Additionally， thoroughly extracting the spectral characteristics of objects， comprehensively understanding the distribution of infrared and hyperspectral image data， and effectively integrating large models with single object tracking are formidable challenges and prominent topics for future research. This paper provides references and insights for future studies.