This paper introduces SKiT, a fast Key information Transformer for phase recognition of videos. Unlike previous methods that rely on complex models to capture long-term temporal information, SKiT accurately recognizes high-level stages of videos using an efficient key pooling operation. This operation records important key information by retaining the maximum value recorded from the beginning up to the current video frame, with a time complexity of ${\mathcal{O}}\left( 1 \right)$. Experimental results on Cholec80 and AutoLaparo surgical datasets demonstrate the ability of our model to recognize phases in an online manner. SKiT achieves higher performance than state-of-the-art methods with an accuracy of 92.5% and 82.9% on Cholec80 and AutoLaparo, respectively, while running the temporal model eight times faster (7ms v.s. 55ms) than LoViT, which uses ProbSparse to capture global information. We highlight that the inference time of SKiT is constant, and independent from the input length, making it a stable choice for keeping a record of important global information, that appears on long surgical videos, essential for phase recognition. To sum up, we propose an effective and efficient model for surgical phase recognition that leverages key global information. This has an intrinsic value when performing this task in an online manner on long surgical videos for stable real-time surgical recognition systems.

A new directional canopy emissivity model (CE-P) based on spectral invariants is proposed in this paper. First, we prove the existence of the spectral invariant properties in the thermal infrared (TIR) band using a Monte Carlo model. Based on it, the equation of the new model is derived from the perspective of absorption. In this expression, single-scattering and multiscattering effects are separated analytically in the TIR band. We find that the overall contribution of multiple scatterings is less than 0.005 when the component emissivities are over 0.90, and the overall contribution decreases with increasing leaf or soil emissivity. Furthermore, the new model can avoid the logical difficulty encountered when using the traditional cavity effect factor to simulate the emissivity of a sparse vegetation canopy. The results of 4SAIL and Discrete Anisotropic Radiative Transfer (DART) are selected to do cross validation. The CE-P can achieve a high accuracy compared with 4SAIL and DART, with an absolute bias less than 0.002 when the leaf (soil) emissivity is equal to 0.98 (0.94). Four widely used analytical models are selected for comparison. The resulting accuracies of these models are ordered from CE-P to REN15, FR97, FR02, and VALOR96 with the most serious error up to 0.002, 0.002, 0.007, 0.013, and 0.014, respectively. Three main conclusions are obtained through the sensitivity analysis: the multiscattering between vegetation and the background can be ignored when the leaf (soil) emissivity is no less than 0.94 (0.90), the second and higher order scattering within the vegetation can also be ignored when the leaf (soil) emissivity is no less than 0.94 (0.90), and the single-scattering effect within the canopy should be considered which can be calculated using three view factors.

The clumping index (CI) is an important factor for deriving forest leaf area index (LAI) at both pixel and point scale. Due to the non-uniform distribution of leaves within trees and random layout of trees under natural conditions, the CI varies with view zenith angle (VZA) as well as the gap fraction. The variation tendency of gap fraction and CI with VZA dominates the potential uncertainty of deriving LAI from directional observations. In this research, the continuous change of gap fraction and CI with VZA was simulated under virtual tree scenarios. Meanwhile, the directional CI was calculated using an algebraic model developed based on previous researches on gap fraction. Both results were validated using ground measurements. The CI discrepancy distributes within ±0.06 coming from the random number distribution restriction and the influence from tree trunk. The analysis using the algebraic model shows that the forest CI reaches an asymptotic level after 55° for forest canopy dense enough. This is an interesting and important phenomenon, which reveal a stable feature of CI to be further used to extract forest information.

A physical model for simulating crop albedo based on the law of energy conservation and spectral invariants was developed. The idea referred to the primary PARAS albedo model for forest. The developed model concerned two key problems for crop monitoring, one is the anisotropy of soil reflectance and the other is the clumping effect of crop leaves at canopy scale, which contributed to the improvement of the model accuracy. The comparison between the model results and those simulated using a Monte Carlo method shows that the calculated clumping index, soil absorptance and canopy albedo all have high accuracy, indicating that the model is able to reflect the interaction mechanism between radiation and the canopy-soil system.