Texture Classification Research Articles

Stochastic analysis of plant available water estimates and soil water balance components simulated by a hydrological model

AbstractThe uncertainty in soil hydraulic parameters is often not taken into account in process‐based hydrological modeling. Performing runs with 104 stochastic parameter realizations, we evaluated the propagation of uncertainty in the Van Genuchten–Mualem (VGM) parameters into estimates of the threshold values of soil water content used to calculate the total and readily available water, and on the long‐term (30 years) simulations of evaporation, transpiration, bottom flux, and runoff by the SWAP hydrological model. The simulated scenarios included weather data from a location in southeast Brazil and seven soils from the same region cropped with maize, comprising a wide range of texture classes. The results showed that uncertainties in VGM parameters affect the estimates of total and readily available water. Water balance components obtained by a deterministic simulation with average VGM parameters did not always agree with the average or median of stochastic simulations, and stochastic simulations including parameter uncertainties should be preferred. Variations in yearly rainfall characteristics were more important for bottom flux and evaporation, while transpiration and runoff were more strongly influenced by the variations in soil hydraulic properties.

Active learning strategies for robotic tactile texture recognition tasks.

Accurate texture classification empowers robots to improve their perception and comprehension of the environment, enabling informed decision-making and appropriate responses to diverse materials and surfaces. Still, there are challenges for texture classification regarding the vast amount of time series data generated from robots' sensors. For instance, robots are anticipated to leverage human feedback during interactions with the environment, particularly in cases of misclassification or uncertainty. With the diversity of objects and textures in daily activities, Active Learning (AL) can be employed to minimize the number of samples the robot needs to request from humans, streamlining the learning process. In the present work, we use AL to select the most informative samples for annotation, thus reducing the human labeling effort required to achieve high performance for classifying textures. We also use a sliding window strategy for extracting features from the sensor's time series used in our experiments. Our multi-class dataset (e.g., 12 textures) challenges traditional AL strategies since standard techniques cannot control the number of instances per class selected to be labeled. Therefore, we propose a novel class-balancing instance selection algorithm that we integrate with standard AL strategies. Moreover, we evaluate the effect of sliding windows of two-time intervals (3 and 6 s) on our AL Strategies. Finally, we analyze in our experiments the performance of AL strategies, with and without the balancing algorithm, regarding f1-score, and positive effects are observed in terms of performance when using our proposed data pipeline. Our results show that the training data can be reduced to 70% using an AL strategy regardless of the machine learning model and reach, and in many cases, surpass a baseline performance. Finally, exploring the textures with a 6-s window achieves the best performance, and using either Extra Trees produces an average f1-score of 90.21% in the texture classification data set.

Identification, Prioritization and Mapping of Major Constraints and Potentials in the Medo Watershed the Case of West Arsi District in Central Rift Valley Area of Ethiopia

The concentration to characterize the biophysical and socio-economic features in the integrated watersheds is to identify and prioritize production constraints for designing appropriate research and development interventions. The objective of this study was to assess the current biophysical of characteristics of Medo watershed, West Arsi Zone of Oromia, Ethiopia. To do so relevant data and tools were used; ArcGIS, Microsoft Excel sheet and fundamental formulae were applied for the analysis are some. The results of the study indicated that the major land use types in the watershed are agricultural land covering 50%, vegetation (shrubs, forests and, plantations) covering 24% and settlement covering 16% out of the total land use. Slope gradient of Medo watershed ranges from zero to more than 20 and the slope gradient of 2-5 and 5-10 cover the greatest in area coverage representing 201 ha and 170 ha respectively. We can also observe that about 8% out of the total area is subjected to severe erosion. Sandy clay loam was the dominant soil textural classes in the surface soils and Phaeozems, Retinols and Andosols are the dominant soil types, which cover 57%, 36% and 7% of the sub-watershed area for lower, middle, and upper slope positions, respectively. Mean annual rainfall of the area ranges from 960.09 mm at lower part of the watershed to the 1304.93 mm at upper part of the watershed. Accordingly, Medo sub-watershed is laid in majority of at sub humid agro ecological zone. An assessment of the trees within the landscape of watershed showed some remnant natural forests and a wide variety of shrubby vegetation is encountered at all landscapes. About 40 percent of the watershed, is under high to extremely severe soil loss values (>45 ton per hectare per year). The baseline study also identified that those natural resources degradation such as land, and or/ soil fertility, reduction, and recent changes in the areas’ weather condition in line of climate change prevailing in current years are few of the many factors that are contributing to the land and crop productivity reductions in the area. Therefore, prioritizing the identified problem and preparing intervention of different technologies and development plans to solve the problems by taking into consideration the existing opportunities of the watershed.

Unsupervised texture classification of 3D X-ray Micro-computed Tomography images

Characterizing rock proprieties is crucial in the oilfield to evaluate hydrocarbon reserves. Several studies showed a high correlation between rock properties and textures. Therefore, we propose integrating texture information in the images to identify precisely the most representative textures in highly heterogeneous rocks to estimate their properties. First, we implemented a steerable pyramid decomposition to extract the texture features. Then, those parameters were used as input for the Self-organizing map to classify the textures. Finally, by applying several models and comparing their results, we suggested the best approach to implement for texture classification.

The ratio of denitrification end-products were influenced by soil pH and clay content across different texture classes in Oklahoma soils

Nitrous oxide (N2O) is a potent greenhouse gas that contributes to stratospheric ozone depletion and global climate change. Soil denitrification has two potential end-products, N2O and dinitrogen (N2), and the ratio of these end-products (N2O:(N2O+N2) or the N2O ratio) is controlled by various factors. This study aims to quantify the influence of soil pH on the ratio of denitrification end-products in Oklahoma soils with different soil textures. Six natural grassland soils encompassing three distinct soil textures were incubated in the laboratory under natural and modified pH with an overall tested pH ranging from 2 to 10. Denitrification end-products were measured in the laboratory using the acetylene inhibition technique and further estimated using a process-based biogeochemical model. Both the laboratory and model results showed that soil pH and texture influenced the ratio of the denitrification end-products. Generally, as soil pH increased the N2O ratio decreased, although both lab and model results indicated that this relationship was not linear. Soil texture may have an indirect effect on the N2O ratio, as two soils of the same texture could have different N2O ratios. However, clay percentage of the soil did show a linear positive correlation with the N2O ratio, suggesting components of soil texture may be more influential than others. Overall, soil pH was a controlling factor in the ratio of denitrification end-products and the newly observed nonlinear relationship warrants further study, particularly when considering its effects in different soil textures.

A novel texture image pyramid based vote strategy in local binary pattern for texture classification

A novel texture image pyramid based vote strategy in local binary pattern for texture classification

Characterization and classification of three major soils at the college of agriculture, Jalingo, Taraba State, Nigeria

Adequate information on land resources have been identified as a pre-requisite for sustainable land management. The present study was carried out to characterize and classify the soils of College of Agriculture, Jalingo, Taraba State, and to suggest soil management recommendations for optimum crop yield. Digital elevation model (DEM) of the study location was obtained and the slope map of the area generated using ArcGIS (ESRI, US) software. This was used to identify the major soils (MU1, MU2 and MU3). Two soil profile pits were sited in each of MU1 and MU2, and one pit was sited in MU3. The soils were described and sampled for laboratory analyses. The results revealed the dominance of transitional horizons namely; AB, BC and CB which also showed raised proportions of clay. The soils were shallow to bedrock and dominated by brown, dark yellowish brown and gray colours (moist) in the surface soils and reddish to yellowish brown in the subsurface horizons. Sand size fraction dominated the particle size distribution and resulted in sandy loam and sandy clay loam textural classes. All the soils had bulk density < 1.50 g cm−3. Soil pH was slightly acid to neutral and values in water were higher than those in CaCl2 in all horizons. Organic carbon, total N, available P and exchangeable bases were low in most of the soils, while base saturation was high. Two of the soils (MU1 and MU2) were classified as Alfisols while MU3 was classified as an Entisol according to the USDA system. According to the World Reference Base for Soil Resources system of classification, MU1 and MU2 were classified as Lixisols and MU3 as a Gleysol. The incorporation of organic manure, cropping across slopes, and practicing afforestation are recommended for soil nutrient management.

Simulasi Model Material Tanah Pada Bidang Gelincir Longsor Dengan Variasi Sudut Kemiringan (Studi Kasus Material Tanah di Daerah Kabupaten Gorontalo Utara)

Landslides are one of the geological disasters that often occur in Indonesia. North Gorontalo is an area where landslides occur every year. Based on data from the 2022 Badan Penanggulangan Bencana Daerah (BPBD), in the last year, there were 8 landslides. Landslides can occur due to several factors, namely; human factors and natural factors. In principle, landslides occur when the restraining force is less than the driving force. In this study a landslide simulation will be carried out using a simulation model made on a lab scale, the size of the simulation model is 40 cm x 30 cm x 25 cm. The simulation was carried out using soil samples taken at the research location. Variables that were observed during the simulation were the effect of the slope angle (o) on the exiting water (ml), the water infiltration time (t), and the effect of the slope (o) on the depth of the slip plane (cm). The soil taken from the research location will be subjected to a sieving analysis test with a hydrometer test to determine the texture class of the soil. From the soil texture test results, some of the research locations have a sand-gravel type of texture. The simulation test carried out also obtained the result that the slope of the slope greatly influences the water infiltration time with the depth of the slip plane. The greater the angle of the slope of the sled that is carried out, the results of the variable being measured will also be directly proportional.

Federated Texture Classification: Implementing Colorectal Histology Image Analysis using Federated Learning

Federated Texture Classification: Implementing Colorectal Histology Image Analysis using Federated Learning

Comment on article "A Simple Classification of Pancreatic Duct Size and Texture Predicts Postoperative Pancreatic Fistula: A Classification of the International Study Group of Pancreatic Surgery".

Comment on article "A Simple Classification of Pancreatic Duct Size and Texture Predicts Postoperative Pancreatic Fistula: A Classification of the International Study Group of Pancreatic Surgery".

Exploring the Physical, Chemical, and Biological Properties of Soils from Different Regions Classified into Different Textural Classes

Exploring the Physical, Chemical, and Biological Properties of Soils from Different Regions Classified into Different Textural Classes

Predictive map of soil texture classes using decision tree model and neural network with features of geomorphology level

This study aims to compare decision tree (DT) and artificial neural network (ANN) models, in addition, the efficiency of geomorphic surface attributes in predicting soil texture classes. The study area is located in the north of Chaharmahal and Bakhtiari province, central west Iran, and covers 6875 ha. Ninety-six pedons were excavated on separated geoforms. Soil samples of top soil (A horizon) were analyzed for clay, sand, and silt contents. Totally 57 auxiliary variables, including the derivatives of digital elevation model (DEM), Landsat 8 images, geomorphic surface map, geology map, and land-use map, were used to predict both soil texture classes and soil particle size fractions. Root-mean-square error (RMSE), R² or the coefficient of determination ( R_square), overall accuracy, and Kappa coefficient were selected as criteria for evaluating model performance. The R-square coefficients of clay, silt, and sand fractions for both models, respectively, were 0.41, 0.25, and 0.63 for ANN and 0.52, 0.62, and 0.75 for DT. According to RMSE, R-square, overall accuracy, and Kapa coefficient of validation data, the DT model produced better prediction fits to the both soil particle-size fraction and soil texture classes and was the most accurate classifier model. The parameters were 0.59, 0.09, 0.66, and 0.24 for ANN and 0.41, 0.75, 0.76, and 0.60 for DT models, respectively. The accuracy of each individual soil texture class was generally dependent upon the number of soil texture observations in each texture class. According to this fact, both models had better prediction for silty clay loam and clay loam texture classes.

Assessment of Soil Fertility Status in Selected Fields under Maize Production in Kongwa District, Dodoma Region, Tanzania

This study was conducted in the Kongwa District to assess the fertility status of soils of the selected fields under maize production to understand fertility variability among soils and recommend appropriate fertilizer rates. The study involved randomly selected 24 maize fields. Composite soil samples were collected in these fields at 0–20 cm deep and characterized for soil fertility status. Results indicated that 48% of the soils were sandy clay loam and 26% were sandy loam. The remaining fields had clay or loamy sand textural classes. The soil pH ranged from extremely acidic (3.52) to moderately alkaline (7.7), organic carbon ranged from very low to medium (0.19-1.60%) and total N were very low to low (0.01-0.15%). Also, results indicated that 42% of soils had P deficiency and 16.7% had inadequate S levels. In addition, 45.8% of the soils had inadequate exchangeable K and exchangeable Mg levels ranged from very low to high (0.29-4.06 cmol(+) kg-1). Exchangeable Ca was low to very high (1.06 to 10.04 cmol(+) kg-1) with favourable base saturation for crop production. The CEC ranged from very low (2.62 cmol(+) kg-1) to medium (18.9 cmol(+) kg-1). Extractable micronutrients such as Cu, Fe, and Mn were adequate but Zn was inadequate in 58% of the soils. Categorizing nutrient status in soils of the study area showed that fertility is poor regarding N, P, K, Zn, Mg, and Ca. Hence, the studied soils need external nutrient inputs and proper management to optimize crop production.

THE DETERMINATION OF THE MOST OPTIMAL METHODS FOR THE IRRIGATION OF CROPS ON DIFFERENT SOIL TYPES

Climate change due to multiple pollution problems (forest clearing, overgrazing, motor vehicle emissions, construction expansion, etc.), has led to increased global average temperature and reduced rainfall. Under these conditions, the mineralization of organic matter is intensified, the activity of soil microorganisms is reduced and implicitly decreases soil trophicity. These aspects further show the need to apply irrigation in the agricultural sector, rationalysing water, managing it and recommending the best methods of water distribution, on soils with different textural classes. In this regard, research was carried out on water permeability, on soils formed in different pedo-climatic conditions. Thus, three soil profiles were opened, in the Tecuci Plain, on a typical arenosoil, in the Dobrogea Plateau, on a calcaric chernozem and in the Gavanu-Burdea Plain on a chromic luvisol. Soil profiles were characterized morphologically and physico-chemically, soil samples were collected both in natural and modified settlement (metal cylinders). The physico-chemical analyses for the three soil types were determined within INCDPA laboratories Bucharest. Based on soil samples collected in metal cylinders, water permeability was determined using the ratio between the thickness of the soil horizon and the time of water infiltration. Depending on the permeability of the soil on the three different soil types, the most suitable and economical watering methods for the current climatic conditions have been established and recommended.

Glaucoma Detection Based on Texture Feature of Neuro Retinal Rim Area in Retinal Fundus Image

One method for detecting glaucoma is by comparing ratios in the area of neuroretinal rim. Comparing area ratios in the neuroretinal rim is difficult for ophthalmologists since it requires high accuracy and is highly dependent on the patient's retinal condition. In this study, we sought to perform neuro retinal rim feature extraction based on histogram and gray level co-occurrence matrix (GLCM) of normal retinal images and glaucoma, automatically distinguish between normal eyes and eyes with glaucoma, and evaluate the method's validity using the measures of accuracy, sensitivity, and specificity We adopted a machine learning approach in conducting automatic feature extraction of the retinal rim through three main stages: 1) image acquisition, 2) pre-processing, and 3) classification. We used a dataset from RIM-ONE for normal eyes images and DRISTHI-GS for glaucoma images.Classification was carried out on 154 images (80 images for glaucoma images and 74 images for normal images). Regarding true positive, false negative, false positive, and true negative, we examined the sensitivity, specificity, and accuracy of automatic extraction and classification. The highest findings are 96.10%, 98.75%, and 93.24%, respectively. This study showed that automatic texture features and classification are possible, accurate and important in detecting glaucoma.

A fast image stitching algorithm based on texture classification and improved SIFT

A fast image stitching algorithm based on texture classification and improved SIFT

Edge‐Preserved Tversky Indexive Hellinger with Deep Perceptive Czekanowski‐Based Image Classification

The texture is identifiable in optical and easy ways. Texture classification is an imperative region in texture analysis, where it gives descriptors for classifying the images. The categorization of normal and abnormal matter by magnetic resonance (MR), computed tomography (CT), and texture images has made noteworthy evolution in modern years. Recently, different novel robust classification techniques have been introduced to classify the different kinds of images for prediction. However, the accuracy of classification was not improved with lesser time. To address these issues, the edge‐preserved Tversky indexive Hellinger and deep perceptive Czekanowski classifier (ETIH‐DPCC) technique is introduced to segment and classify the images with more accuracy. The ETIH‐DPCC technique includes diverse processes namely preprocessing, segmentation, feature extraction, as well as classification. At first, different types of images, such as magnetic resonance imaging, CT, and texture, are used as input. With the acquired input, edge‐preserving normalized adaptive bilateral filtering is employed to carry the image preprocessing. In this stage, the noisy pixels are removed and edges are preserved. Then, the Tversky‐indexed quantile regression is applied to segment the images into diverse texture regions. After that, the feature extraction is performed on the segmented region using Hellinger kernel feature extraction, where a more informative feature for image prediction is extracted. During this process, the irrelevant features are avoided to decrease the dimensionality and feature extraction time. These extracted features are finally classified into positive and negative classes for disease prediction using DPCC. DPCC comprises multiple layers to deeply analyze the association between training and testing features. With this, the prediction accuracy is improved. Experimental outcomes show that the ETIH‐DPCC technique efficiently enhances prediction accuracy and less time compared to conventional methods.

Mutual Dual-Task Generator With Adaptive Attention Fusion for Image Inpainting

Image segmentation can reveal the semantic structure information in an image, which is helpful guidance information for image inpainting. Notably, it can help mitigate the artifacts on the boundaries of different semantic regions during the inpainting process. Existing semantic guidance-based image inpainting provides one-way guidance from the semantic segmentation task to the image inpainting task. There is no feedback from the inpainting results to adjust the guidance process, which causes inferior performance. To tackle this issue, this work proposes mutual dual-task generators to establish the interaction between image segmentation and image inpainting tasks. Thus, semantic segmentation guides image inpainting and also receives feedback from image inpainting. These two processes interact with each other and progressively improve the inpainting quality. The mutual dual-task generator consists of a shared encoder and mutual decoders with the bidirectional Cross-domain Feature DeNormalization (CFDN) module inside, which hierarchically models the Segmentation-guided image Texture (ST) generation and Texture-guided semantic Segmentation (TS) generation. At the end of mutual decoders, an Adaptive Attention Fusion (AAF) module is proposed to augment the texture and semantic class affinity between pixels, further refining the inpainted results. Experimental results demonstrate that the proposed mutual dual-task generator pipeline achieves superior inpainting performances over the state of the arts on three public datasets.

Analysis of the Morphology and Lithology of the Kisangani–Yangambi Section of the Congo River in the Democratic Republic of Congo

Vol. 19, Issue 1, 2024: 59–80 DOI: 10.15170/MG.2024.19.01.0459 Mawa Mukulia, Prométhée1 – Katembo Kasekete, Désiré2 – Makabu Kayembe, Gabriel3 – Muhindo Sahani, Walere 4 – Kahindo Muhongya, Jean-Marie5 Analysis of the Morphology and Lithology of the Kisangani–Yangambi Section of the Congo River in the Democratic Republic of Congo 6 ABSTRACT The Congo River, which is between 900 and 300 metres wide between the city of Kisangani and the town of Yangambi, is characterised by important socio-economic and conservation activities but also by a lack of knowledge on morphological and lithological matters. To fill the gaps, the lithology and the morphological dynamics of the riverbed were studied through a description of islands (between 1984 and 2017) and banks thanks to remote sensing and field work, respectively. With regard to the number of islands, the study revealed a loss of 5.13% and a recovery of 28.21% between 1984 and 2017. The direct factors of this dynamism are the erosion and landslides observed both on the banks and on the islands. The geological survey method allowed the collection of 4 samples of sound indurated rocks and 50 samples of loose rocks on the soil profiles of the islands and banks. The macroscopic description and sieving of the samples with 63 mm and 2 mm mesh size allowed the lithology map to be drawn up. The fine soil particles were analysed in the soil laboratory using the Robinson–Köhn pipette method to determine their texture. The USDA textural classification was used. The results of the analysis showed that sand is the predominant rock element in the area. Such a lithology would increase the sensitivity of the banks to mass movements observed here and there along the Kisangani–Yangambi river section.

Facial Recognition for Criminal Identification using Convolutional Neural Network

The process of identifying and recognising the criminal is the time consuming and difficult task. There are several ways to identify culprits at the crime site, including fingerprinting, DNA matching, and eyewitness testimony. The criminal face identification system will be built on a existing criminal database. The method for identifying a human face using features extrapolated from an image is presented in this study. The technique for identifying a human face using characteristics extrapolated from a picture is presented in this research. It is quite difficult to develop a computer model for recognizing the human face since it is a complicated multidimensional visual representation. The video captured by the camera will be translated into frames as part of the suggested process. To increase detection accuracy, this suggested a Binary Gradient Alignment (BGA) algorithm a description texture classification technique. When a facial feature is detected in an image frame, it undergoes pre-processing to eliminate unnecessary data and reduce unwanted distortions. The real- time processed image is compared to the trained images that have previously been saved in the database. The technology will send an automatic email notice to the police officials if the surveillance camera detects a criminal.