---

Research Interests 

​Machine learning, Computer Vision, Deep learning, Data Science, Remote Sensing. 

---

Research Highlights

dPEN: deep Progressively Expanded Network for mapping of heterogeneous agricultural landscape using WorldView-3 satellite imagery

A novel deep learning method, namely deep progressively expanded network (dPEN), is proposed for mapping nineteen different objects including crop types, weeds and crop residues, in a heterogeneous agricultural field using WorldView-3 (WV-3) imagery.The results demonstrated that: (1) The proposed dPEN allows for building a deeper neural network from multispectral data which was the limitation of many convolutional neural networks; (2) dPEN was able to extract more discriminative features from VNIR and SWIR bands by producing the highest overall accuracy (OA: 86.06%) over competing methods such as support vector machine and random forest; (3) The inclusion of WV-3 SWIR bands greatly improved the classification accuracy; (4) SWIR bands were particularly beneficial to improve the classification accuracy of some individual classes such as weeds, crop residues, and corn and soybean during late developmental stages; (5) The red-edge band (705–745 nm) was identified as the most important band affecting the classification accuracy nearly 10%, whereas the coastal band (400–450 nm) provided the lowest contribution; and (6) SWIR-5 band (2145–2185 nm) contributed most to OA by enhancing it approximately 4% when combined with VNIR bands, while SWIR-1 (1195–1225 nm) yielded the lowest improvement (1.55%) for OA. [Paper]

Suspended sediment concentration estimation from Landsat imagery along the lower Missouri and middle Mississippi Rivers using extreme learning machine

Monitoring and quantifying suspended sediment concentration (SSC) along major fluvial systems such as the Missouri and Mississippi Rivers provide crucial information for biological processes, hydraulic infrastructure, and navigation. Traditional monitoring based on in situ measurements lack the spatial coverage necessary for detailed analysis. This study developed a method for quantifying SSC based on Landsat imagery and corresponding SSC data obtained from United States Geological Survey monitoring stations from 1982 to present. The presented methodology first uses feature fusion based on canonical correlation analysis to extract pertinent spectral information, and then trains a predictive reflectance–SSC model using a feed-forward neural network (FFNN), a cascade forward neural network (CFNN), and an extreme learning machine (ELM). The trained models are then used to predict SSC along the Missouri–Mississippi River system. Results demonstrated that the ELM-based technique generated R2 > 0.9 for Landsat 4–5, Landsat 7, and Landsat 8 sensors and accurately predicted both relatively high and low SSC displaying little to no overfitting. The ELM model was then applied to Landsat images producing quantitative SSC maps. This study demonstrates the benefit of ELM over traditional modeling methods for the prediction of SSC based on satellite data and its potential to improve sediment transport and monitoring along large fluvial systems.. [Paper]

Progressively Expanded Neural Network  (PEN Net) for hyperspectral image classification:
​a new neural network paradigm for remote sensing image analysis

A novel neural network based classification algorithm, named Progressively Expanded Neural Network (PEN Net), is proposed. PEN Net can effectively interpret hyperspectral pixels in nonlinear feature spaces and then determine their categories. Furthermore, a spectral-spatial hyperspectral image (HSI) classification framework is introduced to test the generality and robustness of the PEN Net. Experimental results on four standard hyperspectral data sets illustrate that: (1) PEN Net classifier yields better accuracy and competitive processing speed in HSI classification tasks compared to the state-of-the-art methods; (2) Multi-hidden layer based PEN Net generally provides better performance than single hidden layer one; (3) Combination of spectral and spatial features in the PEN Net classifier can significantly improve the classification accuracy by 6–15% compared to the spectral only based HSI classification. This study implies that the proposed neural network architecture opens a new window for future research and the potential for remote sensing image analysis. [Paper]

Adaptive Trigonometric Transformation Function based Image Enhancement for Unmanned Aerial System Imagery 

Unmanned Aerial System (UAS) based imaging technology has gained great interests in modern photogrammetry and remote sensing. However, due to the limitations of UAS imaging devices, Image Enhancement (IE) has become a necessary process for improving the visual appearance of UAS images. We propose a new adaptive yet highly efficient luminance enhancement method, namely Adaptive Trigonometric Transformation Function (ATTF), for enhancing UAS based images . [Paper] 

​Unmanned Aerial System (UAS)-based phenotyping of soybean using multi-sensor data fusion and extreme learning machine

This study evaluates the power of high spatial resolution RGB, multispectral and thermal data fusion to estimate soybean (Glycine max) biochemical parameters including chlorophyll content and nitrogen concentration, and biophysical parameters including Leaf Area Index (LAI), above ground fresh and dry biomass. ​Results showed that: (1) For biochemical variable estimation, multispectral and thermal data fusion provided the best estimate for nitrogen concentration and chlorophyll (Chl) a content and RGB color information based indices and multispectral data fusion exhibited the largest RMSE 22.6%; the highest accuracy for Chl content estimation was obtained by fusion of information from all three sensors with an RMSE of 11.6%. (2) Among the plant biophysical variables, LAI was best predicted by RGB and thermal data fusion while multispectral and thermal data fusion was found to be best for biomass estimation. (3) For estimation of the above mentioned plant traits of soybean from multi-sensor data fusion, Extreme learning regression yields promising results compared to alternate regression models. This research indicates that fusion of low-cost multiple sensor data within a machine learning framework can provide relatively accurate estimation of plant traits and provide valuable insight for high spatial precision in agriculture and plant stress assessment. ​ [Paper]

​  Volumetric Directional Pattern for Spatial Feature Extraction from Hyperspectral Imagery

In this work, we propose a novel feature extraction method, named Volumetric Directional Pattern, for Hyperspectral Image (HSI) classification. The proposed technique fuses the texture information from three consecutive bands in the input HSI. The extracted local image texture features for each pixel of interest is then fed into an extreme learning machine classifier to assign object category. Experimental results demonstrate that the effectiveness of VDP compared to the state-of-the-arts. [Paper]

Single Image Super Resolution

This work presents a machine learning-based approach to reconstruct a high-resolution (HR) image from a single low- resolution (LR) image. Inspired by the human visual cortex system, which is sensitive to high-frequency (HF) components in an image, we aim to model this concept by training a neural network to estimate the missing HF components that contain structural details. In our method, various directional edge responses at each pixel are considered to obtain more complete HF information and then a regularized extreme learning regression model is trained using a set of LR and HR images. [Paper]

Multiclass Object Detection with Single Query from Hyperspectral Imagery

We propose a deterministic object detection algorithm capable of detecting multiclass objects in hyperspectral imagery (HSI) without any training or preprocessing. The proposed method, which is named class-associative spectral fringe-adjusted joint transform correlation (CSFJTC), is based on joint transform correlation (JTC) between object and nonobject spectral signatures to search for a similar match, which only requires one query (training-free) from the object's spectral signature. The output of CSFJTC yields a pair of sharp correlation peaks for a matched target and negligible or no correlation peaks for a mismatch. Experimental results demonstrate the superiority of the proposed CSFJTC technique compared to the state-of-the-arts. [Paper]

---

Selected Publications 

Book Chapter:

Paheding Sidike, Mohammad Alom and Vasit Sagan, "Robust pattern recognition via joint transform correlation," Sept. 2018. Nova Science Publishers.

Peer-reviewed journals:

11. Maimaitijiang, M., Sagan, V., Paheding Sidike, Maimaitiyiming, M., Hartling, S., Peterson, K.T., Maw, M., Shakoor, N., Mockler, Todd, Fritschi, F. "Vegetation Index Weighted Canopy Volume Model (CVMVI) for soybean biomass estimation from Unmanned Aerial System-based RGB Imagery." ISPRS Journal of Photogrammetry and Remote Sensing [Impact Factor (IF) =5.99],  151:27-41 , May 2019. 

10. Paheding Sidike, Sagan, V., Maimaitijiang, M., Maimaitiyiming, M., Shakoor, N., Burken, J., Mockler, T., Fritschi, F. (2018). dPEN: deep Progressively Expanded Network for mapping of heterogeneous agricultural landscape using WorldView-3 imagery. Remote Sensing of Environment [IF=6.45], vol. 221, 756-772, 2019. [A novel deep learning paradigm and in-depth analysis of WV-3 satellite  imagery for crop monitoring] 

9. Paheding Sidike, V. K. Asari, Vasit  Sagan, "Progressively Expanded Neural Network (PEN Net) for hyperspectral image classification:  a new neural network paradigm for remote sensing image analysis." ISPRS Journal of Photogrammetry and Remote Sensing [IF=5.99], vol. 146, 161-181, Sept. 2018. [A new neural network paradigm for Image Analysis] 

8. Paheding Sidike, V. Sagan, M. Qumsiyeh, M. Maimaitijiang, A. Essa, and V. Asari, “Adaptive Trigonometric Transformation Function with Image Contrast and Color Enhancement: Application to Unmanned Aerial System Imagery,” IEEE Geoscience and Remote Sensing Letters [IF=2.89],  vol. 15, Issue: 3, p.404-408, Mar.  2018.  [Fast and effective illumination enhancement for UAV  data] 

7. M. Maimaitijiang, A. Ghulam, Paheding Sidike, S. Hartling, M. Maimaitiyiming, K. Peterson, E. Shavers, J. Peterson, J. Burken, F. Fritschi, “Unmanned aerial systems based phenotyping of soybean using multi-sensor data fusion and extreme learning machine,” ISPRS Journal of Photogrammetry and Remote Sensing [IF=5.99], vol. 134, p.43-58, 2017.   [Impact of RGB, thermal and Multispectral sensors on soybean phenotyping] 

6. Paheding Sidike, Vijayan K. Asari, and Mohammad S. Alam, “Multiclass object detection with single query in hyperspectral imagery using class-associative spectral fringe-adjusted joint transform correlation,” IEEE Transactions on Geoscience and Remote Sensing (TGRS) [IF=4.66], vol.54, issue 2, p.1196-1208, Feb. 2016. [The first formulation of deterministic multiclass object detection via joint correlation analysis] 

5. ALmabrok Essa, Paheding Sidike, and Vijayan K. Asari, “Volumetric directional pattern for hyperspectral image classification,” IEEE Geoscience and Remote Sensing Letters [IF=2.89], vol. 14, no. 7, p. 1-5, 2017. [New texture feature through spectral direction] 

4. Evan Krieger, Paheding Sidike, Theus Aspiras, and Vijayan K. Asari, “Deterministic object tracking using Gaussian ringlet and directional edge features,” Optics & Laser Technology [IF=2.50], vol. 95, p. 133-146, May 2017.   [Rotation Invariant object tracking] 

3. Paheding Sidike, Evan Krieger, Zahangir Alom, Vijayan K. Asari and Tarek Taha, “A fast single image super-resolution via directional edge guided regularized extreme learning regression,” Signal, Image and Video Processing [IF=1.64], vol. 11, p. 961–968, 2017.   [Neural network based near real-time single image super-resolution] 

2. Zahangir Alom, Paheding Sidike, Tarek Taha and Vijayan K. Asari, “State preserving extreme learning machine: a monotonically increasing learning approach,” Neural Processing Letters [IF=1.78], vol. 45, issue 2, p. 703–725, Sept. 2016.  [Very simple but effective ELM method]

1. Paheding Sidike, M. S. Alam, “Logarithmic fringe-adjusted joint transform correlation,” Optical Engineering [IF=0.99], 52(10), p. 103108, 2013.