珠海一号-2和3高光谱卫星分别于2018年4月26日和2019年9月19日发射成功。数据辐射质量评价是遥感数据应用的基础之一，针对珠海一号高光谱数据，基于辐射精度、清晰度、信噪比和信息熵四个客观指标，对珠海高光谱 L1B级数据辐射质量进行评价，并与GF-5号高光谱遥感数据相同谱段（440-1000nm）数据辐射质量进行对比。结果表明：GF-5号高光谱数据的辐射精度和清晰度均优于珠海高光谱数据，并且珠海高光谱数据的清晰度为GF-5号数据清晰度的54.5%左右；在信息熵方面，两者能力近似，均在在6-10之间；在信噪比方面，珠海高光谱数据的信息熵为GF-5号数据信息熵的86.5%左右。因此，珠海和GF-5号高光谱数据在一定程度上可以补充使用，同时珠海高光谱数据可通过提高量化级数、降低光谱分辨率和优化传感器探元响应的方式提高数据辐射质量。

Objective: The Orbita Hyper Spectral (OHS)-2 and 3 satellites were successfully launched on April 26, 2018 and September 19, 2019, respectively. The basis of remote sensing data applications is the data quality evaluation. However, there has been no systematic or previous evaluation or research on the radiation quality evaluation of the OHS at present. Method: There are 32 bands in the products of OHS. It will consume a lot of manpower, material resources and time if the products are evaluated by subjective evaluation method. Therefore, this study mainly focused on the objective evaluation method to evaluate the radiation quality of OHS level 1B images. The radiation quality of oHS-2 and OHS-3 were evaluated by objective evaluation method in the regions covered by representative features and similar time. Based on the four objective indexes, namely radiation accuracy, image definition (EVA), signal to noise ratio (SNR) and entropy, the radiation quality of OHS level 1B images were evaluated and the radiation quality of OHS and GF-5 (440-1000 nm) images were compared. Result: The results showed that the radiation quality and EVA of GF-5 were higher than OHS, and the EVA of OHS was about 54.5% of that in GF-5; and the ability of the entropy was similar: the entropy of OHS was about 91.5% of that in GF -5, and the value range of entropy is 6-10. At the same time, the signal to noise ratio (SNR) of OHS is about 86.5% of that of GF-5. Thus, the OHS and GF-5 data could be supplemented. Meanwhile, the OHS could improve the data radiation quality by improving the quantitative series of spectral resolution, reducing the spectral resolution and optimizing the sensor response. Conclusion: This study provides data quality reference for the applications of OHS images. The radiation quality of OHS-2C and OHS-3B was evaluated by four objective indexes: radiation accuracy, EVA, SNR and entropy. At the same time, the radiation quality of GF-5 was also compared with OHS. Although the radiation quality of OHS is lower than that of GF-5 due to the restriction of spectral resolution, and the SNR and EVA of GF-5 data are obviously better than that of OHS, but the entropy of OHS and GF-5 is very similar. Due to the high revisit cycle of OHS (6 days for single-star network, 2 days for 4-star network) and high spatial resolution (10m), OHS images can complement GF-5 image to a certain degree in remote sensing applications. In the future, we will study the spectral quality and atmospheric correction of OHS in terms of quantitative remote sensing and water quality monitoring.