Application of the XBoost Regressor for an A Priori Prediction of UAV Image Quality

In general, the quality of imagery from Unmanned Aerial Vehicles (UAVs) is evaluated after the flight, and then a decision is made on the further value and use of the acquired data. In this paper, an a priori (preflight) image quality prediction methodology is proposed to estimate the preflight image quality and to avoid unfavourable flights, which is extremely important from a time and cost management point of view. The XBoost Regressor model and cross-validation were used for machine learning of the model and image quality prediction. The model was learned on a rich database of real-world images acquired from UAVs under conditions varying in both sensor type, UAV type, exposure parameters, weather, topography, and land cover. Radiometric quality indices (SNR, Entropy, PIQE, NIQE, BRISQUE, and NRPBM) were calculated for each image to train and test the model and to assess the accuracy of image quality prediction. Different variants of preflight parameter knowledge were considered in the study. The proposed methodology offers the possibility of predicting image quality with high accuracy. The correlation coefficient between the actual and predicted image quality, depending on the number of parameters known a priori, ranged from 0.90 to 0.96. The methodology was designed for data acquired from a UAV. Similar prediction accuracy is expected for other low-altitude or close-range photogrammetric data.

[1]  Cezary Specht,et al.  Using UAV Photogrammetry to Analyse Changes in the Coastal Zone Based on the Sopot Tombolo (Salient) Measurement Project , 2020, Sensors.

[2]  J. J. Muñoz-Pérez,et al.  Beach Leveling Using a Remotely Piloted Aircraft System (RPAS): Problems and Solutions , 2020, Journal of Marine Science and Engineering.

[3]  Guido Morgenthal,et al.  Quality Assessment of Unmanned Aerial Vehicle (UAV) Based Visual Inspection of Structures , 2014 .

[4]  Rafal Mantiuk,et al.  Comparison of Four Subjective Methods for Image Quality Assessment , 2012, Comput. Graph. Forum.

[5]  Qianqing Qin,et al.  Blind UAV Images Deblurring Based on Discriminative Networks , 2018, Sensors.

[6]  Damian Wierzbicki,et al.  Methodology of improvement of radiometric quality of images acquired from low altitudes , 2016 .

[7]  Xiaoling Chen,et al.  Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes , 2006 .

[8]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[9]  Navid Razmjooy,et al.  Deep learning as an alternative to super-resolution imaging in UAV systems , 2020 .

[10]  Damian Wierzbicki,et al.  Influence of the Sun Position and Platform Orientation on the Quality of Imagery Obtained from Unmanned Aerial Vehicles , 2020, Remote. Sens..

[11]  Qing Liu,et al.  A physical-based atmospheric correction algorithm of unmanned aerial vehicles images and its utility analysis , 2017 .

[12]  Shahram Shirani,et al.  Subjective and Objective Quality Assessment of Image: A Survey , 2014, ArXiv.

[13]  Xuan Wang,et al.  Airborne Infrared and Visible Image Fusion Combined with Region Segmentation , 2017, Sensors.

[14]  Lu Yuan,et al.  Image deblurring with blurred/noisy image pairs , 2007, ACM Trans. Graph..

[15]  Damian Wierzbicki,et al.  A Method for Dehazing Images Obtained from Low Altitudes during High-Pressure Fronts , 2019, Remote. Sens..

[16]  Hongguang Li,et al.  Haze removal for UAV reconnaissance images using layered scattering model , 2016 .

[17]  Ewelina Rupnik,et al.  A two-step approach for the correction of rolling shutter distortion in UAV photogrammetry , 2020 .

[18]  Michael J. Starek,et al.  Deep Learning-Based Single Image Super-Resolution: An Investigation for Dense Scene Reconstruction with UAS Photogrammetry , 2020, Remote. Sens..

[19]  Pawel Burdziakowski,et al.  Polymodal Method of Improving the Quality of Photogrammetric Images and Models , 2021, Energies.

[20]  Li Zhao,et al.  Haze concentration adaptive network for image dehazing , 2021, Neurocomputing.

[21]  Patricia Ladret,et al.  The blur effect: perception and estimation with a new no-reference perceptual blur metric , 2007, Electronic Imaging.

[22]  Ming-Hsuan Yang,et al.  Single Image Dehazing via Multi-scale Convolutional Neural Networks with Holistic Edges , 2019, International Journal of Computer Vision.

[23]  Fatih Erden,et al.  RF-based low-SNR classification of UAVs using convolutional neural networks , 2020, ITU Journal on Future and Evolving Technologies.

[24]  Fourry Handoko,et al.  Assessing a 35mm Fixed-Lens Sony Alpha-5100 Intrinsic Parameters Prior to, During, and Post UAV Flight Mission , 2019 .

[25]  Mohammad Shorif Uddin,et al.  Image Quality Assessment through FSIM, SSIM, MSE and PSNR—A Comparative Study , 2019, Journal of Computer and Communications.

[26]  Zhang Qi-shan.  A UAV Reconnaissance Image Quality Assessing Method , 2007 .

[27]  Pawel Burdziakowski,et al.  A Novel Method for the Deblurring of Photogrammetric Images Using Conditional Generative Adversarial Networks , 2020, Remote. Sens..

[28]  Joonki Paik,et al.  Wavelength-Adaptive Dehazing Using Histogram Merging-Based Classification for UAV Images , 2015, Sensors.

[29]  Eija Honkavaara,et al.  Radiometric Correction of Close-Range Spectral Image Blocks Captured Using an Unmanned Aerial Vehicle with a Radiometric Block Adjustment , 2018, Remote. Sens..

[30]  Syed Muhammad Anwar,et al.  Natural scene statistics model independent no-reference image quality assessment using patch based discrete cosine transform , 2020, Multimedia Tools and Applications.

[31]  M. Zontul,et al.  The prediction of the ZnNi thickness and Ni % of ZnNi alloy electroplating using a machine learning method , 2021 .

[32]  Jim H. Chandler,et al.  Automatic detection of blurred images in UAV image sets , 2016 .

[33]  Damian Wierzbicki,et al.  Influence of Lower Atmosphere on the Radiometric Quality of Unmanned Aerial Vehicle Imagery , 2019, Remote. Sens..

[34]  A. Stoffel,et al.  Photogrammetric reconstruction of homogenous snow surfaces in alpine terrain applying near-infrared UAS imagery , 2017 .

[35]  Shannon R. Clemens,et al.  PROCEDURES FOR CORRECTING DIGITAL CAMERA IMAGERY ACQUIRED BY THE AGGIEAIR REMOTE SENSING PLATFORM , 2012 .

[36]  Damian Wierzbicki,et al.  Quality Assessment of the Bidirectional Reflectance Distribution Function for NIR Imagery Sequences from UAV , 2018, Remote. Sens..