Texture Classification Research Articles

Renyi entropy analysis of a deep convolutional representation for texture recognition

Despite the recent success of convolutional neural networks in computer vision in general, texture images still pose an important challenge to those models, especially when dealing with textures “in the wild”. In this context, deep learning models can benefit from the introduction of features long known to be useful for texture modeling. And that is the case of entropy, a measure of texture regularity that has played an important role in classical computer vision. Based on this observation, here we propose an alternative analysis over deep convolutional neural features based on entropy for texture representation and, particularly, texture classification. More precisely, we couple a module to the convolutional backbone that locally computes the Renyi entropy of the latent representation at multiple levels. The rationale for using Renyi entropy is essentially two-fold: (1) It connects the concept of entropy with multifractal theory, another well explored measure especially in domains of application where we seek some physical interpretation of the descriptors, e.g., in medicine, biology, and others; (2) It has an extra degree of freedom (α parameter) that can be fine-tuned. The main contribution of this study is the development of a strategy that can improve the performance of convolutional neural networks in texture recognition tasks, adding low computational cost. The effectiveness of our method is verified in texture classification of benchmark datasets, as well as in a practical task of plant species identification. Our method achieves a competitive accuracy of 84.5% in the classification of KTH-TIPS-2b and 80.9% in FMD, which are two challenging benchmark databases. Besides, an accuracy of 91.6% is obtained in the plant identification application, to our knowledge outperforming all the results previously published on this task. In both scenarios, the proposed descriptors outperform several approaches from the state-of-the-art, confirming the method potential as a rich and robust solution for texture analysis in general. The results also suggest that information-theoretical measures like entropy can be a reliable source of information to compose a precise and robust latent representation of texture images.

On evaluating the hypothesis of shape similarity between soil particle-size distribution and water retention function

Two pedotransfer functions (PTFs) are available in the literature enabling the soil water retention function (WRF) to be estimated from knowledge of the soil particle-size distribution (PSD), oven-dry soil bulk density (b), and saturated soil water content (s): i) the Arya and Heitman model (PTF-AH) and ii) the Mohammadi and Vanclooster model (PTF-MV). These physico-empirical PTFs rely on the hypothesis of shape similarity between PSD and WRF, and do not require the calibration of the input parameters. In the first stage, twenty-seven PSD models were evaluated using 4,128 soil samples collected in Campania (southern Italy). These models were ranked according to the root mean square residuals (RMSR), corrected Akaike Information Criterion (AICc), and adjusted coefficient of determination (R2adj). In the second stage, three subsets of PSD and WRF data (DS-1, DS-2, and DS-3), comprising 282 soil samples, were used to evaluate the two PTFs using the best three PSD models selected in the first stage. The hypothesis of shape similarity was assumed as acceptable only when the RMSR value was lower than the field standard deviation of the WRFs (*), which is viewed as a tolerance threshold and computed from the physically-based scaling approach proposed by Kosugi and Hopmans (1998). In the first study area (DS-1), characterized by a fairly uniform, loamy textured volcanic soil, the PTF-AH outperformed the PTF-MV and both PTFs provided reasonable performance within the acceptance threshold (i.e., RMSR < *). In the other two heterogeneous field sites (DS-2 and DS-3, characterized by soil textural classes that span from clay and clay-loam to loam and even sandy-loam soils), the PTF-MV (with 3% to 6% RMSR surpassing *) outperformed the PTF-AH (with 8% to 30% RMSR surpassing *) and the majority of RMSR values were larger than those obtained in the original studies. The mean relative error (MRE) revealed that the PTF-MV systematically underestimates the measured WRFs, whereas the PTF-AH provided negative MRE values indicating an overall overestimation. The outcomes of our study provide a critical evaluation when using calibration-free PTFs to predict WRFs over large areas.

Fertility Status of Soils of Patuakhali Science and Technology University in Bangladesh

Soil is a heterogeneous body and different soil have different fertilizer requirement for crop production depending on fertility status. This research work was conducted to evaluate the fertility status of the soils of the research field of Patuakhali Science and Technology University (PSTU). The soils of PSTU were collected and analyzed for the determination of following soil parameters viz. particle size distribution, texture, pH, organic matter, total nitrogen, available phosphorus, available sulphur and exchangeable potassium. The results showed that the particle size distribution of the collected soils varied significantly with each other. A remarkable variation was observed in textural classes of the soils. The textural classes of the soils were sandy loam, clay and clay loam. The soils were neutral to slightly alkaline in reaction and the pH ranged from 7.01 to 7.73. The organic matter content was very low and ranged from 0.50 to 0.96%. The soils were found to be deficient in total nitrogen with a range of 0.06 to 0.11%. The available phosphorus, available sulphur and exchangeable potassium content of collected soils ranged from 8.76 to 15.90 ppm, 1.04 to 3.06 ppm and 0.12 to 0.26 meq/100 g soil, respectively. The fertility status of the soils was very low to medium level due to very low organic matter content.

Formation of magnesium‐clay in a lacustrine microbialite‐bearing carbonate deposit, Eocene Green River Formation, Sanpete County, Utah

AbstractAuthigenic magnesium‐clays have been observed and documented in the Green River Formation, specifically as ooidal stevensite grains. Magnesian clays are a valuable proxy for reconstructing shallow‐water, saline–alkaline lake palaeoenvironments due to their responses to chemical changes. Magnesium‐rich clay minerals are relatively common components in modern and ancient lake systems. A rare interaction of lacustrine magnesian clays, microbialites, carbonates and volcanoclastic deposits of the Eocene Green River Formation crop out in Sanpete Valley, Utah, USA. The characterization of this interaction, Mg‐rich clay and carbonate minerals genesis, the environmental controls on Mg‐rich clay minerals distribution and accumulation is still poorly understood. This study has identified various species of Mg‐clay minerals (stevensite, mixed‐layer kerolite–stevensite and sepiolite), along with calcite and dolomite, found in four Mg‐clay‐bearing lithofacies: (i) Mg‐clay stromatolites; (ii) Mg‐clay arenites; (iii) Mg‐claystone; and (iv) intraclastic Mg‐clay hybrid arenites and conglomerates. Magnesium‐clay bearing rocks from the Sanpete Valley area were deposited in a south‐western, shallow embayment of Eocene Lake Uinta along the margin of the Sevier fold and thrust belt. Sanpete Valley's Mg‐clay‐bearing section shows syngenetic mineralogical paragenesis with neoformation of magnesian clays accompanied by calcite and silica. Regional chronostratigraphic correlation of Sanpete Valley's Mg‐clay‐bearing section with the Uinta Basin was made by zircon U–Pb dating (laser ablation–inductively coupled plasma–mass spectrometry) of two tuff beds. Stevensite and kerolite are precipitated under very specific conditions. This study presents a depositional model, considering evidence from the authigenic mineralogy, facies, fossil evidence and basin context. All of these findings allow a comprehensive analysis of the lithostratigraphic and chemostratigraphic evolution of this isolated bay system, as well as texture classification. This study provides analogue for detailed correlation and comparison to other basins containing similar Mg‐clay and microbialite‐bearing deposits.

FabricNET: A Microscopic Image Dataset of Woven Fabrics for Predicting Texture and Weaving Parameters through Machine Learning

This research presents an approach aimed at enhancing texture recognition and weaving parameter estimation in the textile industry to align with sustainability goals and improve product quality. By utilizing low-cost handheld microscopy and machine learning, this method offers the potential for more precise production outcomes. In this study, textile images were manually labeled for texture, specific mass, weft, and warp parameters, followed by the extraction of various texture features, resulting in a comprehensive dataset comprising four hundred and fifty-eight inputs and four outputs. Prominent machine learning algorithms, including XGBoost, RF, and MLP, were applied, resulting in noteworthy achievements. Specifically, XGBoost demonstrated an impressive texture classification accuracy of 0.987, while RF yielded the lowest MAE (5.121 g/cm) in specific mass prediction. Additionally, weft and warp estimations displayed superior accuracy compared to manual measurements. This research emphasizes the crucial role of AI in improving efficiency and sustainability within the textile industry, potentially reducing resource wastage, enhancing worker safety, and increasing productivity. These advancements hold the promise of significant positive environmental and social impacts, marking a substantial step forward in the industry’s pursuit of its objectives.

A Completed Multiply Threshold Encoding Pattern for Texture Classification

The binary pattern family has drawn wide attention for texture representation due to its promising performance and simple operation. However, most binary pattern methods focus on local neighborhoods but ignore center pixels. Even if some studies introduce the center based sub-pattern to provide complementary information, extant center based sub-patterns are much weaker than other local neighborhood based sub-patterns. This severe unbalance limits the classification performance of fusion features significantly. To alleviate this problem, this paper designs a multiply threshold center pattern (MTCP) to provide a more discriminative and complementary local texture representation with a compact form. First, a multiply thresholds encoding strategy is designed to encode the center pixel that generates three 1-bit binary patterns. Second, it adopts a compact multi-pattern encoding strategy to combine them into the 3-bit MTCP. Furthermore, this paper proposes a completed multiply threshold encoding pattern by fusing the MTCP, local sign pattern, and local magnitude pattern. Comprehensive experimental evaluations on three popular texture classification benchmarks confirm that the completed multiply threshold encoding pattern achieves superior texture classification performance.

Urban Flooding Vulnerability Analysis Using Weighted Linear Method With Geospatial Information System

This study was carried out to investigate the causative factors responsible for flooding in Etsako east, Nigeria. The study applied weighted linear method and geospatial information technology to obtain relevant data which were subjected to analysis. Sentinel 2 imagery of the area was downloaded from the United States Geological Survey (USGS) website and analyzed to produced digital elevation model, rainfall distribution, soil texture classification, topographic wetness index (TWI), hillside slope, land use/cover contour and slope maps of the area. Saaty weighted scale (1 – 5) was used to assign importance to the factors selected. The results presented using relevant maps revealed that, TWI (14%), precipitation (14%), distance to river (12%), elevation (11%) and slope (10%) are the major factors responsible for flooding in the area. Using built up areas only (46.57 km2 ); 38.6% of it is within the risk range of High to very High flood vulnerability, moderate vulnerability zone (17.97 km2 ) occupied 27.28%. Low vulnerability zone occupied 22.05%,(10.27 km2 ) while no vulnerability zone occupied 12.05% (5.61 km2). Communities in the high flood vulnerable region includes Iddo, Iyukwe, Iyogeh, Imiegb and Opepwe; flood control measures and early warning systems should be developed to mitigate the impact of flooding in the area.

Multi-scale LBP fusion with the contours from deep CellNNs for texture classification

In texture classification, local binary pattern (LBP) is currently one of the most widely-concerned feature encoding models. Most existing LBP-based texture classification methods are usually limited to single-kind texture features. In fact, an across-domain fusion of LBP features with other features, such as image contours, could be another potential path to promote texture classification. To enhance the feature modeling ability of LBP-based methods, this paper firstly designs a Cellular Neural Network (CellNN) with recurrent convolutions, initially trained by a simplified simulated-annealing algorithm, to extract informative image contours. For better reliability, a new three-channel contour extractor of deep CellNNs (i.e., dCellNNs) is proposed. This extractor contains the initially-trained CellNNs of more than three layers, and it is further optimized by fine-tuning parameters. Moreover, a new weighted-base algorithm is designed to fulfill the fusion of the multi-scale texture features by LBPs and the contour features by dCellNNs to enhance feature representation. Finally, these enhanced features are concatenated together to generate the final multi-scale features of given texture image. On texture datasets KTH, Brodatz, OTC12 and UIUC, experiment results verify that the across-domain fusion of multi-scale LBPs and dCellNNs is efficient in capturing & enhancing texture features. With moderate feature dimensionality and computational costs, it could improve texture classification, acquiring an obvious accuracy increase on previous state-of-the-art ones, e.g., a rise of 2.58% on KTH-TIPS2b, a rise of 3.11% on Brodatz, a rise of 0.71% on OTC12 and a rise of 0.42% on UIUC.

Combining contour-based and region-based in image segmentation

Background: This paper presents an optimized clustering approach applied to image segmentation. Accurate image segmentation impacts many fields like medical, machine vision, object detection. Applications involve tumor detection, face detection and recognition, and video surveillance. Methods: The developed approach is based on obtaining an optimum number of clusters and regions of an image. We combined Region-based and contour-based approaches. Initial rough regions are obtained using edge detection. We have used Gabor wavelets for texture classification and spatial resolutions. Color frequencies are also used to determine the number of clusters of the Fuzzy c-means (FCM) algorithm which gave an optimum number of clusters or regions. Results: We have compared our approach with other similar wavelet and clustering techniques. Our algorithm gave better values for segmentation metrics like SNR, PSNR, and MCC. Conclusions: Optimizing the number of clusters or regions has a significant effect on the performance of the image segmentation techniques. This will result in better detection and localization of the segmentation-based application.

A New Texture Spectrum Based on Parallel Encoded Texture Unit and Its Application on Image Classification: A Potential Prospect for Vision Sensing.

In industrial applications based on texture classification, efficient and fast classifiers are extremely useful for quality control of industrial processes. The classifier of texture images has to satisfy two requirements: It must be efficient and fast. In this work, a texture unit is coded in parallel, and using observation windows larger than 3×3, a new texture spectrum called Texture Spectrum based on the Parallel Encoded Texture Unit (TS_PETU) is proposed, calculated, and used as a characteristic vector in a multi-class classifier, and then two image databases are classified. The first database contains images from the company Interceramic®® and the images were acquired under controlled conditions, and the second database contains tree stems and the images were acquired in natural environments. Based on our experimental results, the TS_PETU satisfied both requirements (efficiency and speed), was developed for binary images, and had high efficiency, and its compute time could be reduced by applying parallel coding concepts. The classification efficiency increased by using larger observational windows, and this one was selected based on the window size. Since the TS_PETU had high efficiency for Interceramic®® tile classification, we consider that the proposed technique has significant industrial applications.

Effects of land use and slope position on selected soil physicochemical properties in Tekorsh Sub-Watershed, East Gojjam Zone, Ethiopia

Abstract In the Tekorsh Sub-Watershed, East Gojjam Zone, Ethiopia, researchers investigated the impact of land use patterns and slope position on selected soil physico-chemical parameters. The study area was arbitrarily divided into three slope positions (higher, middle, and lower), two land uses types (grazing and cultivated land), and two soil depths (0–20 and 20–40 cm) with three replications, based on the in situ field survey. For laboratory analysis, a total of 36 composite samples were obtained. Sand, clay, and silt fraction were highly significantly (P ≤ 0.05) affected by the interaction effect of the three factors. Clay and clay loam were the textural classes of the soil in the study area. The interaction effects of the three factors were highly significant (P ≤ 0.001) affected bulk density (D b), total porosity (TP), organic carbon (OC), available phosphorus (AP), exchangeable (Mg2+, Ca2+, and acidity), cation exchange capacity (CEC), and micronutrients (Fe, Mn, Cu, and Zn). The soils were medium to high (1.22–1.44 g/cm3) in D b, very high (45.66–63.13%) in TP, medium to high (15.72–19.56% v/v) in available water holding capacity, low to medium (1.37–2.91%) in OC, very low (1.65–7.68 mg/kg) in AP, high (4.62–5.36 cmol(+)/kg) in exchangeable Mg2+, very high in CEC (43.60–51.06 cmol(+)/kg), Fe (25.20–52.91 mg/kg), Mn (37.29–105.55 mg/kg), Cu (4.04–7.87 kg/kg), and Zn (0.83 2.53 kg/kg). In general, it was discovered that the majority of the assessed soil properties were better in grazing land than in soils utilized for cultivated land uses, and that the lower slope position was preferable to the upper and middle ones.

Thermal Mapping from Point Clouds to 3D Building Model Facades

Thermal inspection of buildings regarding efficient energy use is an increasing need in today’s energy-demanding world. This paper proposes a framework for mapping temperature attributes from thermal point clouds onto building facades. The goal is to generate thermal textures for three-dimensional (3D) analysis. Classical texture generation methods project facade images directly onto a 3D building model. Due to the limited level of detail of these models, projection errors occur. Therefore, we use point clouds from mobile laser scanning extended by intensities extracted from thermal infrared (TIR) image sequences. We are not using 3D reconstructed point clouds because of the limited geometric density and accuracy of TIR images, which can lead to poor 3D reconstruction. We project these thermal point clouds onto facades using a mapping algorithm. The algorithm uses a nearest neighbor search to find an optimal nearest point with three different approaches: “Minimize angle to normal”, “Minimize perpendicular distance to normal”, and “Minimize only distance”. Instead of interpolation, nearest neighbor is used because it retains the original temperature values. The thermal intensities of the optimal nearest points are weighted by resolution layers and mapped to the facade. The approach “Minimize perpendicular distance to normal” yields the finest texture resolution at a reasonable processing time. The accuracy of the generated texture is evaluated based on estimating the shift of the window corner points from a ground truth texture. A performance metric root-mean-square deviation (RMSD) value that measures this shift is calculated. In terms of accuracy, the nearest neighbor method outperformed bilinear interpolation and an existing TIR image-based texturing method.

Soil macroporosity modulates the extent of negative microbial associations during organic substance decomposition

Microbial species interactions are expected to influence the community-level properties, such as the production of extracellular enzymes and the degradation of organic substances. This work examined how microbial diversity, composition and the overall sign of microbial associations were altered with soil texture and structure following the amendment of organic substances. Two sets of microcosms (1:100 and 1:1) of a 4 × 3 factorial design were constructed, with four artificial textural classes (a sandy loam, two loams, and a (silty) clay loam) and three organics (TSB, tryptic soy broth; CA, a mixture of cellulose and humic/fulvic acids; BS, barley straw). As the ‘microbial inoculant’, an agricultural soil was added to the 1:100 and 1:1 microcosms at 1% and 50%, respectively. A few of microbial taxa were specifically enriched after soil addition of TSB, CA, or BS, but distributions across textural classes were inconsistent between microcosms or between organic amendments. Regardless, top abundant bacterial and fungal OTUs were overall negatively associated, suggesting that microbial competition for the shared resource dominated the decomposition of both simple and complex organics. Microbial associations were also modified by soil pore size distribution (PSD), being fewer negative (or more positive) in soils of greater macroporosity than in soils of lower macroporosity. The PSD-based differences in microbial associations were coordinated with PSD-based differences in the activities of exoglucanase and β-glucosidase in TSB-amended soils or soil respiration characteristics in CA-amended soils. Our results provide new insight into how soil structure regulates microbial interactions and, accordingly, the degradation of organic matter.

Genetic Features of Soils on Dense Parent Rocks of the Donetsk Ridge

The present study investigates morphological, physical and chemical properties of Chernozems at dense parent rocks. The research was carried out on the protected areas native soils of the north-west of the Rostov region. This territory is representing eastern slopes of the Donetsk Ridge, which predetermines the dense rocks release to the surface. It has been found that morphological, physical and chemical properties of Chernozems on dense rocks have emphasized their originality and uniqueness, and at the same time clearly have indicate their classification as Chernozem order. In comparison with the zonal soil – Calcic Chernozem – Chernozems on dense rocks are characterized by a reduced thickness of humus-accumulative layers, an increased of humus content, the presence of gravel in profile bottom half, a weak expression of calcareous accumulations, features of the particle-size distribution of soil components, even if it belongs to the same soil texture class. Chernozems on dense carbonate rocks either do not have accumulation or are characterized by a weak manifestation of secondary carbonates (pseudomycelia and nodules). This fact has differed such soil from the same order Chrernozems on fine-grained rocks (Calcic Chernozems) for which it is diagnostic criteria. Therefore, the mention of the suborder of these soils is incorrect in the designation. In the soil covering of the spurs of the Donetsk Ridge, dark humus soils are also occure, previously classified as incomplete developed chernozems, differing from chernozems by the absence of medium part of soil profiles.

Soil Attributes and Their Interrelationships with Resistance to Root Penetration and Water Infiltration in Areas with Different Land Uses in the Apodi Plateau, Semiarid Region of Brazil

Studies on soils and their interrelationships with land use in the context of the semi-arid region of Brazil are still scarce, even though they have the potential to assist in understanding the use and management of soil and agricultural crops. From this perspective, this study investigated four land uses in different locations of the Apodi Plateau, an elevated area in semi-arid region of northeastern Brazil. The different soils were analyzed for their resistance to root penetration, water infiltration, inorganic fractions, soil density, total porosity, potential of hydrogen, electrical conductivity, total organic carbon, potential acidity, and sum of bases. The soil resistance to root penetration and water infiltration were determined in the field. The results obtained were interpreted using multivariate and geostatistical analysis. The resistance data were subjected to the Shapiro–Wilk test at 5% of probability and expressed in maps, whereas infiltration data curves were constructed to estimate the amount of infiltrated water at the different time intervals. The textural classification was an important factor for the analysis of soil resistance to root penetration (Q) and the infiltration rate, being evidenced in the cluster analysis and allowing the formation of two groups, one for the surface layers of the areas and another for the subsurface layers, with the inorganic sand and clay fractions standing out with the greatest dissimilarity. The establishment of conservation practices for soil management is suggested to correct the pore space problems and the degradation of agroecosystems in areas with soils whose conditions are similar to the ones of this study.

Texture classification and in situ cation measurements by SEM-EDX provide detailed quantitative insights into cell wall cation interaction changes during ageing, soaking, and cooking of red kidney beans

Hard-to-cook (HTC) is a textural defect that delays the softening of common bean seeds during cooking. While this defect is commonly associated with conventionally stored beans, soaking/cooking of beans in CaCl2 solutions or sodium acetate buffer can also prolong the cooking time of beans due to formation of Ca2+ crosslinked pectin retarding bean softening during cooking. In this study, the role of the cell wall-bound Mg2+/Ca2+ content and the degree of pectin methyl esterification (DM) was quantified, as important factors for bean texture-related changes stipulated in the pectin-cation-phytate hypothesis, the most plausible hypothesis of HTC development. Evaluation of texture changes during cooking of conventionally aged beans (35 °C and 83% RH for up to 20 weeks), beans soaked/cooked in CaCl2 solutions (0.01 to 0.1 M) or soaked in 0.1 M sodium acetate buffer (pH 4.4) revealed large bean-to-bean variations. Therefore a texture-based classification approach was used to better capture the relation between texture characteristics and cell wall polymer, in particular pectin, related changes. While cell wall-bound Ca2+ and pectin DM did not change/were not related to the texture variation during cooking of fresh beans, increased cell wall-bound Ca2+ and decreased pectin DM were associated with prolonged conventional storage of beans and their texture changes during subsequent cooking (due to pectin cross linking, retarding its solubilization during cooking). Exogenously added Ca2+ from pre-treating beans in CaCl2 solutions promoted to a great extent the cell wall-bound Ca2+ during soaking but even more so during cooking, complementing the harder texture associated with these beans during cooking (compared to conventionally stored and fresh beans). Similarly, free Ca2+ endogenously generated by phytase-catalysed phytate hydrolysis (beans treated by acetate buffer) promoted crosslinking of pectin by Ca2+ (cell wall-bound Ca2+), delaying softening of beans during cooking.

Fast CU Partition Algorithm for Intra Frame Coding Based on Joint Texture Classification and CNN.

High-efficiency video coding (HEVC/H.265) is one of the most widely used video coding standards. HEVC introduces a quad-tree coding unit (CU) partition structure to improve video compression efficiency. The determination of the optimal CU partition is achieved through the brute-force search rate-distortion optimization method, which may result in high encoding complexity and hardware implementation challenges. To address this problem, this paper proposes a method that combines convolutional neural networks (CNN) with joint texture recognition to reduce encoding complexity. First, a classification decision method based on the global and local texture features of the CU is proposed, efficiently dividing the CU into smooth and complex texture regions. Second, for the CUs in smooth texture regions, the partition is determined by terminating early. For the CUs in complex texture regions, a proposed CNN is used for predictive partitioning, thus avoiding the traditional recursive approach. Finally, combined with texture classification, the proposed CNN achieves a good balance between the coding complexity and the coding performance. The experimental results demonstrate that the proposed algorithm reduces computational complexity by 61.23%, while only increasing BD-BR by 1.86% and decreasing BD-PSNR by just 0.09 dB.

Red pyterlite from Virolahti in southeastern Finland: a unique heritage stone with a classic rapakivi texture applied in historic and modern architecture

Red pyterlite from Virolahti in southeastern Finland: a unique heritage stone with a classic rapakivi texture applied in historic and modern architecture

Retina disease prediction using modified convolutional neural network based on Inception‐ResNet model with support vector machine classifier

AbstractArtificial intelligence and deep learning have aided ocular disease through experiments including automatic illness recognition from images of the iris, fundus, or retina. Automated diagnosis systems (ADSs) provide services for the benefit of humanity and are essential in the early detection of harmful diseases. In fact, early detection is essential to avoid total blindness. In real life, several diagnostic tests such as visual ocular tonometry, retinal exam, and acuity test are performed, but they are conclusively time demanding and stressful for the patient. To consume time and detect the retinal disease earlier, an efficient prediction method is designed. In this proposed model, the first process is data collection that consists of a retinal disease dataset for testing and training. The second process is pre‐processing, which executes image resizing and noise filter for feature extraction. The third step is feature extraction, which extracts the image's form, size, color, and texture for classification with CNN based on Inception‐ResNet V2. The classification process is done by using the SVM with the extracted features. The prediction of diseases is classified such as normal, cataract, glaucoma, and retinal disease. The suggested model's performance is assessed using performance indicators such as accuracy, error, sensitivity, precision, and so forth. The suggested model's accuracy, error, sensitivity, and precision are 0.96, 0.962, 0.964, and 0.04, respectively, higher than existing techniques such as VGG16, Mobilenet V1, ResNet, and AlexNet. Thus, the proposed model instantly predicts retinal disease.

Field-scale heterogeneity overrides management impacts following conversion to no-till within an arable system

Crop establishment in no-till arable systems benefits from soil health conducive for growth. Combined with the incorporation of crop residues and manures, no-till can influence soil organic carbon (SOC) and organic matter (SOM) dynamics, crop productivity and nutrient cycling. These processes are shaped by spatial and temporal factors including associated microbial activity. This study aimed to investigate the interaction between microbial and soil physicochemical properties during the transition from full-inversion to no-till soil management. Assessments were conducted over a two-year period and included a combination of soil microbial assays (microbial biomass carbon and nitrogen with physicochemical analyses, SOC, SOM, textural class, pH, gravimetric water content, and macronutrients). Two experiments were established within the same four-hectare field: one on a relatively level area (Experiment-1) and another on a slope (Experiment-2). Experiment-1 treatments consisted of Farmyard Manure (FYM and N-fertiliser), Green Manure (GM, Raphanus sativus and Vicia sp. mix) and Standard Practice (SP = Control, N-fertiliser only). Experiment-2 was a repeat of Experiment-1, but without the FYM treatment. Soil was sampled twice per crop season, in Spring and Autumn, in Experiment-1, and in Autumn only in Experiment-2. The results were influenced by spatial (i.e. where the same was collected from) and temporal (i.e. the time at which the same was collected) variations that were not always linked to management practices. This study demonstrated that the quantification of SOC and SOM were poor predictors of change in management practices over two years, while microbial biomass responded quickly to the incorporation of FYM. SOC and SOM were affected by soil texture, but not significantly by inputs, and were associated with extractable Ca2+ and total-N. Diachronic studies increase our understanding of biological and physicochemical dynamics in response to short-term change in soil management practices. This study emphasises the impact of soil texture within a single heterogenous field, and how it affects management outcomes. It highlights the importance of considering spatial differences to develop effective and sustainable agricultural solutions.