Hafiz Muhammad Shahzad Asif’s research while affiliated with University of Engineering and Technology and other places

Remote sensing is increasingly being used in agriculture and smart farming. Crop cover identification is a major challenge that is useful in the identification of a particular crop at scale. Various studies are conducted to address this challenge using remote sensing and machine learning techniques, but there is still room for improvement in predictive performance. This study has addressed the problem by incorporating multiple modalities for classification modeling. The study has used high-resolution satellite images to perform classification using convolutional neural networks. Densenet201 provided the highest classification accuracy among five candidate architectures. NDVI is calculated from medium-resolution images to be used as a feature that is combined with weather records containing temperature, rainfall, and humidity. A classification ensemble is trained on these features to perform crop classification. Five classification models are trained to select the best classification model. The classification models are evaluated with a train/test split as the models are trained using 60% of the data whereas 20% data is used for validation and 20% for testing. A meta-learner is trained on classification probabilities of both of these classifiers and final class label is obtained. Among the four meta-learners, support vector machines provided best learning and yielded a classification accuracy of 98.83% and an f1-score of 98.78%. High classification performance of the proposed approach indicate that multimodality and meta-learners are useful choices to improve the predictive performance for crop cover identification and can be successfully used for this task.

The accumulation of salt through natural causes and human artifice, such as saline inundation or mineral weathering, is marked as salinization, but the hindrance toward spatial mapping of soil salinity has somewhat remained a consistent riddle despite decades of efforts. The purpose of the current study is the spatial mapping of soil salinity in Kot Addu (situated in the south of the Punjab province, Pakistan) using Landsat 8 data in five advanced machine learning regression models, i.e., Random Forest Regressor, AdaBoost Regressor, Decision Tree Regressor, Partial Least Squares Regression and Ridge Regressor. For this purpose, spectral data were obtained between 20 and 27 of January 2017 and a field survey was carried out to gather a total of fifty-five soil samples. To evaluate and compare the model’s performances, the coefficient of determination (R2), Mean Squared Error (MSE), Mean Absolute Error (MAE) and the Root-Mean-Squared Error (RMSE) were used. Spectral data of band values, salinity indices and vegetation indices were employed to study the salinity of soil. The results revealed that the Random Forest Regressor outperformed the other models in terms of prediction, achieving an R2 of 0.94, MAE of 1.42 dS/m, MSE of 3.58 dS/m and RMSE of 1.89 dS/m when using the Differential Vegetation Index (DVI). Alternatively, when using the Soil Adjusted Vegetation Index (SAVI), the Random Forest Regressor achieved an R2 of 0.93, MAE of 1.46 dS/m, MSE of 3.90 dS/m and RMSE of 1.97 dS/m. Hence, remote sensing technology with machine learning models is an efficient method for the assessment of soil salinity at local scales. This study will contribute to mitigating osmotic stress and minimizing the risk of soil erosion by providing early warnings regarding soil salinity. Additionally, it will assist agriculture officers in estimating soil salinity levels within a shorter time frame and at a reduced cost, enabling effective resource allocation.

Digital images contain a lot of redundancies, therefore, compression techniques are applied to reduce the image size without loss of reasonable image quality. Same become more prominent in the case of videos which contains image sequences and higher compression ratios are achieved in low throughput networks. Assessment of quality of images in such scenarios has become of particular interest. Subjective evaluation in most of the scenarios is infeasible so objective evaluation is preferred. Among the three objective quality measures, full-reference and reduced-reference methods require an original image in some form to calculate the image quality which is unfeasible in scenarios such as broadcasting, acquisition or enhancement. Therefore, a no-reference Perceptual Image Quality Index (PIQI) is proposed in this paper to assess the quality of digital images which calculates luminance and gradient statistics along with mean subtracted contrast normalized products in multiple scales and color spaces. These extracted features are provided to a stacked ensemble of Gaussian Process Regression (GPR) to perform the perceptual quality evaluation. The performance of the PIQI is checked on six benchmark databases and compared with twelve state-of-the-art methods and competitive results are achieved. The comparison is made based on RMSE, Pearson and Spearman correlation coefficients between ground truth and predicted quality scores. The scores of 0.0552, 0.9802 and 0.9776 are achieved respectively for these metrics on CSIQ database. Two cross-dataset evaluation experiments are performed to check the generalization of PIQI.

Blind image quality assessment is a challenging task particularly due to the unavailability of reference information. Training a deep neural network requires a large amount of training data which is not readily available for image quality. Transfer learning is usually opted to overcome this limitation and different deep architectures are used for this purpose as they learn features differently. After extensive experiments, we have designed a deep architecture containing two CNN architectures as its sub-units. Moreover, a self-collected image database BIQ2021 is proposed with 12,000 images having natural distortions. The self-collected database is subjectively scored and is used for model training and validation. It is demonstrated that synthetic distortion databases cannot provide generalization beyond the distortion types used in the database and they are not ideal candidates for general-purpose image quality assessment. Moreover, a large-scale database of 18.75 million images with synthetic distortions is used to pretrain the model and then retrain it on benchmark databases for evaluation. Experiments are conducted on six benchmark databases three of which are synthetic distortion databases (LIVE, CSIQ and TID2013) and three are natural distortion databases (LIVE Challenge Database, CID2013 and KonIQ-10 k). The proposed approach has provided a Pearson correlation coefficient of 0.8992, 0.8472 and 0.9452 subsequently and Spearman correlation coefficient of 0.8863, 0.8408 and 0.9421. Moreover, the performance is demonstrated using perceptually weighted rank correlation to indicate the perceptual superiority of the proposed approach. Multiple experiments are conducted to validate the generalization performance of the proposed model by training on different subsets of the databases and validating on the test subset of BIQ2021 database.

Achieving food security has become a major challenge for society. Crop yield estimation is essential for crop monitoring to ensure food security. Manual crop yield estimation is cumbersome and inaccurate and becomes infeasible when scaled up. Machine learning algorithms trained using remotely sensed data have played a vital role in estimating the yield of different crops. Furthermore, to enrich the data provided to a machine learning algorithm, multiple modalities can be combined to improve the predictive performance of these algorithms. In this research, we propose to combine data from multiple modalities, i.e., agrometeorological and remote sensing data, to predict the tea yield at the farm level. The dataset employed in this study is acquired from tea fields of the National Tea and High-Value Crop Research Institute (NTHRI), Mansehra, Pakistan. The remote sensing data of the Landsat-8 satellite is converted to farm-level NDVI statistics through geocoding. Before being used for regression modeling, the final dataset is subjected to some further preprocessing steps, including the selection of features and the optimization of feature sets. This preprocessed data is used to train the three classes of machine learning regression algorithms. Conventional regression algorithms, including Decision Trees, Multilayer Perceptron (MLP), Support Vector Regression (SVR), Gaussian Process Regression (GPR), and Multiple Linear Regression applied with and without interaction terms and stepwise feature inclusion with various kernels. Moreover, the following three variants of the ensemble learning methods have also been applied: random forest, gradient boosting, and XgBoost. Finally, this study proposed a neural architecture for tea yield estimation using Landsat imagery. This deep neural network is built using neural architecture search via Bayesian optimization and have three hidden layers, which can perform complex non-linear modeling. Experimental evaluation is performed through 10-fold cross-validation, and the proposed Deep neural network regression model provided the best predictive performance. The model provided a coefficient of determination (R-squared) of 0.99 with a Mean Square Error (MSE) of 108.17 kg/ha, Root Mean Square Error (RMSE) of 10.87 kg/ha, Mean Absolute Error (MAE) of 2.26 kg/ha and Mean Absolute Percentage Error (MAPE) of 2.92.

Crop diseases are a major threat to food security and their rapid identification is important to prevent yield loss. Swift identification of these diseases are difficult due to the lack of necessary infrastructure. Recent advances in computer vision and increasing penetration of smartphones have paved the way for smartphone-assisted disease identification. Most of the plant diseases leave particular artifacts on the foliar structure of the plant. This study was conducted in 2020 at Department of Computer Science and Engineering, University of Engineering and Technology, Lahore, Pakistan to check leaf-based plant disease identification. This study provided a deep neural network-based solution to foliar disease identification and incorporated image quality assessment to select the image of the required quality to perform identification and named it Agricultural Pathologist (Agro Path). The captured image by a novice photographer may contain noise, lack of structure, and blur which result in a failed or inaccurate diagnosis. Moreover, AgroPath model had 99.42% accuracy for foliar disease identification. The proposed addition can be especially useful for application of foliar disease identification in the field of agriculture.

Blind image quality assessment is a challenging task particularly due to the unavailability of reference information. Training a deep neural network requires a large amount of training data which is not readily available for image quality. Transfer learning is usually opted to overcome this limitation and different deep architectures are used for this purpose as they learn features differently. After extensive experiments, we have designed a deep architecture containing two CNN architectures as its sub-units. Moreover, a self-collected image database BIQ2021 is proposed with 12,000 images having natural distortions. The self-collected database is subjectively scored and is used for model training and validation. It is demonstrated that synthetic distortion databases cannot provide generalization beyond the distortion types used in the database and they are not ideal candidates for general-purpose image quality assessment. Moreover, a large-scale database of 18.75 million images with synthetic distortions is used to pretrain the model and then retrain it on benchmark databases for evaluation. Experiments are conducted on six benchmark databases three of which are synthetic distortion databases (LIVE, CSIQ and TID2013) and three are natural distortion databases (LIVE Challenge Database, CID2013 and KonIQ-10 k). The proposed approach has provided a Pearson correlation coefficient of 0.8992, 0.8472 and 0.9452 subsequently and Spearman correlation coefficient of 0.8863, 0.8408 and 0.9421. Moreover, the performance is demonstrated using perceptually weighted rank correlation to indicatethe perceptual superiority of the proposed approach. Multiple experiments are conducted to validate the generalization performance of the proposed model by training on different subsets of the databases and validating on the test subset of BIQ2021 database.

Tea (Camellia sinensis L.) is one of the most highly consumed beverages globally after water. Several countries import large quantities of tea from other countries to meet domestic needs. Therefore, accurate and timely prediction of tea yield is critical. The previous studies used statistical, deep learning, and machine learning techniques for tea yield prediction, but crop simulation models have not yet been used. However, the calibration of a simulation model for tea yield prediction and the comparison of these approaches is needed regarding the different data types. This research study aims to provide a comparative study of the methods for tea yield prediction using the Food and Agriculture Organization (FAO) of the United Nations AquaCrop simulation model and machine learning techniques. We employed weather, soil, crop, and agro-management data from 2016 to 2019 acquired from tea fields of the National Tea and High-Value Crop Research Institute (NTHRI), Pakistan, to calibrate the AquaCrop simulation model and to train regression algorithms. We achieved a mean absolute error (MAE) of 0.45 t/ha, a mean squared error (MSE) of 0.23 t/ha, and a root mean square error (RMSE) of 0.48 t/ha in the calibration of the AquaCrop model and, out of the ten regression models, we achieved the lowest MAE of 0.093 t/ha, MSE of 0.015 t/ha, and RMSE of 0.120 t/ha using 10-fold cross-validation and MAE of 0.123 t/ha, MSE of 0.024 t/ha, and RMSE of 0.154 t/ha using the XGBoost regressor with train test split. We concluded that the machine learning regression algorithm performed better in yield prediction using fewer data than the simulation model. This study provides a technique to improve tea yield prediction by combining different data sources using a crop simulation model and machine learning algorithms.

Digital images contain a lot of redundancies, therefore, compressions are applied to reduce the image size without the loss of reasonable image quality. The same become more prominent in the case of videos that contains image sequences and higher compression ratios are achieved in low throughput networks. Assessment of the quality of images in such scenarios becomes of particular interest. Subjective evaluation in most of the scenarios becomes infeasible so objective evaluation is preferred. Among the three objective quality measures, full-reference and reduced-reference methods require an original image in some form to calculate the quality score which is not feasible in scenarios such as broadcasting or IP video. Therefore, a non-reference quality metric is proposed to assess the quality of digital images which calculates luminance and multiscale gradient statistics along with mean subtracted contrast normalized products as features to train a Feedforward Neural Network with Scaled Conjugate Gradient. The trained network has provided good regression and R2 measures and further testing on LIVE Image Quality Assessment database release-2 has shown promising results. Pearson, Kendall, and Spearman's correlation are calculated between predicted and actual quality scores and their results are comparable to the state-of-the-art systems. Moreover, the proposed metric is computationally faster than its counterparts and can be used for the quality assessment of image sequences.

Identification of plant disease is usually done through visual inspection or during laboratory examination which causes delays resulting in yield loss by the time identification is complete. On the other hand, complex deep learning models perform the task with reasonable performance but due to their large size and high computational requirements, they are not suited to mobile and handheld devices. Our proposed approach contributes automated identification of plant diseases which follows a sequence of steps involving pre-processing, segmentation of diseased leaf area, calculation of features based on the Gray-Level Co-occurrence Matrix (GLCM), feature selection and classification. In this study, six color features and twenty-two texture features have been calculated. Support vector machines is used to perform one-vs-one classification of plant disease. The proposed model of disease identification provides an accuracy of 98.79% with a standard deviation of 0.57 on tenfold cross-validation. The accuracy on a self-collected dataset is 82.47% for disease identification and 91.40% for healthy and diseased classification. The reported performance measures are better or comparable to the existing approaches and highest among the feature-based methods, presenting it as the most suitable method to automated leaf-based plant disease identification. This prototype system can be extended by adding more disease categories or targeting specific crop or disease categories.