Algorithm evaluation study reveals reliable image quality assessment in unmanned aerial vehicle hyperspectral images, suggesting improved automated monitoring across diverse flight conditions.
Currently, unmanned aerial vehicle hyperspectral images (UAV-HSIs) lack quick, objective, and comprehensive image quality assessment methods. Therefore, a multi-feature-based fuzzy comprehensive evaluation method was proposed in this letter to comprehensively evaluate UAV-HSI quality. To characterize the hyperspectral quality comprehensively, we selected four radiometric features, three spatial features, and two spectral features to construct an indicator set. After analyzing the statistical distribution of the above features of the 23 UAV-HSIs, a fuzzy evaluation threshold table and membership functions were established. To determine the optimal feature weights, the weights obtained using the three methods were tested on a test set of UAV-HSIs constructed with different degrees of noise and blur. The test results showed that the combined weight based on a combination of subjective and objective weight is more robust. The experiment for comprehensive quality assessment of different distortion types and flight heights was done with UAV-HSIs. The results indicated that the comprehensive quality score was in good agreement with the subjective assessment and the objective fact. This comprehensive quality evaluation method can be effectively used for blurred, noisy, overexposed, and different-height UAV-HSIs.
No takes yet. Share an insight, caveat, or question.
Tian et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: