Search

The crack detection of concrete bridges is an important link in the safety evaluation of bridge structures, and the rapid and accurate identification and detection of bridge cracks is a prerequisite for ensuring the safety and long-term stable use of bridges. To solve the incomplete crack detection and segmentation caused by the complex background and small proportion in the actual bridge crack images, this paper proposes a coarse–fine combined bridge crack detection method of "double detection + single segmentation" based on deep learning. To validate the effect and practicality of fine crack detection, images of old civil bridges and viaduct bridges against a complex background and images of a bridge crack against a simple background are used as datasets. You Only Look Once V5(x) (YOLOV5(x)) was preferred as the object detection network model (ODNM) to perform initial and fine detection of bridge cracks, respectively. Using U-Net as the optimal semantic segmentation network model (SSNM), the crack detection results are accurately segmented for fine crack detection. The test results showed that the initial crack detection using YOLOV5(x) was more comprehensive and preserved the original shape of bridge cracks. Second, based on the initial detection, YOLOV5(x) was adopted for fine crack detection, which can determine the location and shape of cracks more carefully and accurately. Finally, the U-Net model was used to segment the accurately detected cracks and achieved a maximum accuracy (AC) value of 98.37%. The experiment verifies the effectiveness and accuracy of this method, which not only provides a faster and more accurate method for fine detection of bridge cracks but also provides technical support for future automated detection and preventive maintenance of bridge structures and has practical value for bridge crack detection engineering. [ABSTRACT FROM AUTHOR]

A high-quality dataset is a basic requirement to ensure the training quality and prediction accuracy of a deep learning network model (DLNM). To explore the influence of label image accuracy on the performance of a concrete crack segmentation network model in a semantic segmentation dataset, this study uses three labelling strategies, namely pixel-level fine labelling, outer contour widening labelling and topological structure widening labelling, respectively, to generate crack label images and construct three sets of crack semantic segmentation datasets with different accuracy. Four semantic segmentation network models (SSNMs), U-Net, High-Resolution Net (HRNet)V2, Pyramid Scene Parsing Network (PSPNet) and DeepLabV3+, were used for learning and training. The results show that the datasets constructed from the crack label images with pix-el-level fine labelling are more conducive to improving the accuracy of the network model for crack image segmentation. The U-Net had the best performance among the four SSNMs. The Mean Intersection over Union (MIoU), Mean Pixel Accuracy (MPA) and Accuracy reached 85.47%, 90.86% and 98.66%, respectively. The average difference between the quantized width of the crack image segmentation obtained by U-Net and the real crack width was 0.734 pixels, the maximum difference was 1.997 pixels, and the minimum difference was 0.141 pixels. Therefore, to improve the segmentation accuracy of crack images, the pixel-level fine labelling strategy and U-Net are the best choices. [ABSTRACT FROM AUTHOR]

The crack detection of concrete bridges is an important link in the safety evaluation of bridge structures, and the rapid and accurate identification and detection of bridge cracks is a prerequisite for ensuring the safety and long-term stable use of bridges. To solve the incomplete crack detection and segmentation caused by the complex background and small proportion in the actual bridge crack images, this paper proposes a coarse–fine combined bridge crack detection method of “double detection + single segmentation” based on deep learning. To validate the effect and practicality of fine crack detection, images of old civil bridges and viaduct bridges against a complex background and images of a bridge crack against a simple background are used as datasets. You Only Look Once V5(x) (YOLOV5(x)) was preferred as the object detection network model (ODNM) to perform initial and fine detection of bridge cracks, respectively. Using U-Net as the optimal semantic segmentation network model (SSNM), the crack detection results are accurately segmented for fine crack detection. The test results showed that the initial crack detection using YOLOV5(x) was more comprehensive and preserved the original shape of bridge cracks. Second, based on the initial detection, YOLOV5(x) was adopted for fine crack detection, which can determine the location and shape of cracks more carefully and accurately. Finally, the U-Net model was used to segment the accurately detected cracks and achieved a maximum accuracy (AC) value of 98.37%. The experiment verifies the effectiveness and accuracy of this method, which not only provides a faster and more accurate method for fine detection of bridge cracks but also provides technical support for future automated detection and preventive maintenance of bridge structures and has practical value for bridge crack detection engineering.

A high-quality dataset is a basic requirement to ensure the training quality and prediction accuracy of a deep learning network model (DLNM). To explore the influence of label image accuracy on the performance of a concrete crack segmentation network model in a semantic segmentation dataset, this study uses three labelling strategies, namely pixel-level fine labelling, outer contour widening labelling and topological structure widening labelling, respectively, to generate crack label images and construct three sets of crack semantic segmentation datasets with different accuracy. Four semantic segmentation network models (SSNMs), U-Net, High-Resolution Net (HRNet)V2, Pyramid Scene Parsing Network (PSPNet) and DeepLabV3+, were used for learning and training. The results show that the datasets constructed from the crack label images with pix-el-level fine labelling are more conducive to improving the accuracy of the network model for crack image segmentation. The U-Net had the best performance among the four SSNMs. The Mean Intersection over Union (MIoU), Mean Pixel Accuracy (MPA) and Accuracy reached 85.47%, 90.86% and 98.66%, respectively. The average difference between the quantized width of the crack image segmentation obtained by U-Net and the real crack width was 0.734 pixels, the maximum difference was 1.997 pixels, and the minimum difference was 0.141 pixels. Therefore, to improve the segmentation accuracy of crack images, the pixel-level fine labelling strategy and U-Net are the best choices.

Wheat is the third most important cereal crop in Nepal after rice and maize both in area and production, but its productivity of 2.3 tonne ha−1 is very less compared to other developed countries (6 tonne ha−1 for Switzerland and China) in the world. The main cause of low wheat yield in Nepal is the improper and inadequate use of fertilizer devoid of site specific nutrient management practices. Therefore, a farmers' field experiment was conducted during November 2015 to April 2016 to rectify the best fertilizer management options at two sites of Damak and Gauradaha in Jhapa district in eastern-Terai of Nepal using Nutrient Expert®-Wheat model. The research was accomplished in Randomized Complete Block Design with 2 treatments and 20 replications, considering farmers' field as replication. Two treatments included in the experimentation were NE® (Nutrient Expert Recommendation) and FFP (Farmer's Fertilizer Practices). The statistical result revealed the highly significant difference in terms of number of effective tiller m−2, plant height, filled grain per spike, spike length, grain, straw and biological yields and harvest index. The highest yield (4.71 tonne ha−1) was obtained from NE field than FFP (2.99 tonne ha−1). On an average, NE based practices produced 58 % higher yield in comparison to FFP. NE based treatment produced significantly higher biomass yield, yield attributes and cost-benefit ratio than FFP treatments. Field experiment validation confirmed that the Nutrient Expert® Wheat model could be used as the most adoptable and practical precision decision support system tool to make a more authentic fertilizer recommendation in eastern-Terai of Nepal.

Irrigated rice (Oryza sativa L.) and wheat (Triticum aestivum L.) yield increases in Punjab, India, have slowed in recent years. Future yield increases may occur in smaller increments through fine-tuning of crop management mainly by accounting for the large spatial and temporal variation in soil characteristics. On-farm experiments were conducted from 2002-03 to 2004-05 at 56 sites in six key irrigated rice-wheat domains of Punjab to evaluate an approach for site-specific nutrient management (SSNM). Site-specific N–P–K applications were calculated by accounting for the indigenous nutrient supply, yield targets, and nutrient demand as a function of the interactions between N, P, and K. The performance of SSNM was tested for two rice and two wheat crops. Compared with the current farmers’ fertilizer practice (FFP), average grain yields increased from 5.1 to 6.0 and 4.2 to 4.8 Mg/ha for rice and wheat crops, respectively, while plant N, P, and K accumulations increased by 12 to 20%. Improved timing and/or splitting of fertilizer N increased N recovery efficiency from 0.20 and 0.17 kg/kg in FFP plots to 0.30 and 0.27 kg/kg in SSNM plots in rice and wheat crops, respectively. The agronomic N use efficiency was 63 to 83% greater with SSNM than with FFP. The gross return above fertilizer cost (GRF) was 13-14% greater with SSNM than with FFP. Future SSNM research must build on the present approach to develop a more practical way for achieving similar benefits across large areas without site-specific modeling and with minimum crop monitoring.

Mục tiêu của nghiên cứu nhằm đánh giá phương pháp xác định lượng phân N, P và K cần bón cho bắp lai dựa trên năng suất bắp lai trên đất phù sa An Phú – An Giang. Thí nghiệm được thực hiện vào 2 vụ Đông Xuân 2014 -2015 và 2015 – 2016 với sáu nghiệm thức (i) bón NPKCaMg (200N); (ii) bón khuyết N; (iii) bón khuyết P; (iv) bón khuyết K; (v) bón NPKCaMg (160N); (vi) thực tế bón phân của nông dân (FFP). Trên nguyên lý SSNM, phương pháp xác định nhu cầu phân N dựa vào hiệu quả thu hồi (REN) với liều lượng phân đáp ứng nhu cầu năng suất bắp lai thực tế địa phương so với bón theo phương pháp hiệu quả nông học (AEN). Nhu cầu phân P và K được xác định dựa vào lượng phân được loại bỏ bằng hạt và lượng phân tăng theo đáp ứng năng suất mục tiêu. Đất phù sa An Phú – An Giang trên cùng một năng suất đạt được (11-12 tấn/ha) nhu cầu bón NPK trên đất có bao đê cao hơn đất không bao đê. Khả năng cung cấp N từ đất đạt từ 45-50%, đối với P và K khả năng cung cấp từ đất >80%, khả năng cung cấp dưỡng chất NPK từ đất theo thứ tự K>P>N. Đất bao đê có khả năng cung cấp nguồn dinh dưỡng NPK từ đất thấp hơn so với đất không bao đê, khả năng cung cấp dưỡng chất NPK từ đất theo thứ tự 51-80-91%; 54-86-91%.

Understanding, predicting, and managing the spatiotemporal complexity of stream thermal regimes requires monitoring strategies designed specifically to make inference about spatiotemporal variability on the whole stream network. Moreover, monitoring can be tailored to capture particular facets of this complex thermal landscape that may be important indicators for species and life stages of management concern. We applied spatial stream network models (SSNMs) to an empirical dataset of water temperature from the Snoqualmie River watershed, WA, and use results to provide guidance with respect to necessary sample size, location of new sites, and selection of a modeling approach. As expected, increasing the number of monitoring stations improved both predictive precision and the ability to estimate covariates of stream temperature; however, even relatively small numbers of monitoring stations, n = 20, did an adequate job when well-distributed and when used to build models with only a few covariates. In general, winter data were easier to model and, across seasons, mean temperatures were easier to model than summer maximums, winter minimums, or variance. Adding new sites was advantageous but we did not observe major differences in model performance for particular new site locations. Adding sites from parts of the river network with thermal regimes which differed from the rest of the network, and which were therefore highly influential, improved nearby predictions but reduced model-estimated precision of predictions in the rest of the network. Lastly, using models which accounted for the network-based spatial correlation between observations made it much more likely that estimated prediction confidence intervals covered the true parameter; the exact form of the spatial correlation made little difference. By incorporating spatial structure between observations, SSNMs are particularly valuable for accurate estimation of prediction uncertainty at unmeasured locations. Based on our results, we make the following suggestions for designing water temperature monitoring arrays: (1) make use of pilot data when possible; (2) maintain a distribution of monitors across the stream network (i.e., over space and across the full range of covariates); (3) maintain multiple spatial clusters for more accurately estimating correlation of nearby sites; (4) if sites are to be added, prioritize capturing a range of covariates over adding new tributaries; (5) maintain a sensor array in winter; and (6) expect reduced accuracy and precision when predicting metrics other than means. [ABSTRACT FROM AUTHOR]

On farm trials were conducted for two year to evaluate the effect of soil test based site specific nutrient management (SSNM) vis-a-vis other farmers practice on rice (Oryza sativa L.) and wheat (Triticum aestivum L.) productivity in villages of Ghaziabad (Uttar Pradesh) and Faridabad (Haryana). Results revealed that the SSNM led to significant increase in grain yield of rice and wheat crops in both the years as compared to the state recommended NPK fertilizer treatment and farmers practices. On an average, SSNM increased the grain yield of rice crop by 19.9% and 14.9% at Partapur and Badarpur said, respectively, over farmers practice. The corresponding increases in grain yield of wheat at Kumbhawas and Badarpur said were 13.9% and 15.4 %. The maximum net profits of Rupees 9 401/ha and Rupees 12 100/ha were obtained from rice under SSNM at Badarpur Said and Partapur, respectively. In wheat crop, the highest net profits of Rupees 6 346/ ha at Badarpur said and Rupees 4 346/ha at Kumbhawas were recorded under SSNM. The lowest yield and net profits from both the crops were obtained with farmers practices. The physicochemical properties (pH, EC and organic carbon) do not get deteriorated due to various nutrient management practices. Available major (NPK), secondary (S) and micronutrients (Zn and B) exhibited significant differences due to treatment variations. SSNM proved significantly superior in respect of buildup of these nutrients in soil.

Elephant foot yam [Amorphophallus paeoniifolius (Dennst.) Nicolson] is an important tuber crop grown in many parts of India for its starchy corm and the average yield of this crop is far below the potential productivity of 100 tonnes/ha. One of the reasons attributed to the lower yield is the imbalanced application of essential plant nutrients. Site specific nutrient management (SSNM) is the application of plant nutrients based on the soil and crop need, yield target and developed with the aid of models such as QUEFTS. This paper gives the result of four year study conducted to calibrate and validate the QUEFTS model for elephant foot yam cultivation and the development of SSNM zonation maps and secondary and micronutrient fortified customised fertilizers for the crop in major growing environments of India. Data collected from different field experiments conducted in major elephant foot yam-production regions of India during 1968 to 2011 were used to calibrate the model. The derived parameters of minimum and maximum accumulation of N (130 and 460), P (900 and 2100) and K (100 and 170) are proposed as standard borderline values in the QUEFTS model for elephant foot yam. A linear increase in corm yield was suggested by the model with N, P and K uptakes of 3.97, 0.71, and 7.05 kg N, P and K/1 000 kg corm. The average NPK ratio in total plant dry matter was 5.56:1:9.88. Based on these results, the model was calibrated using historical data as well as by conducting field experiments. It can be observed that fertilizer best management practices by SSNM resulted in an average actual corm yield of 33.45 tonnes/ha, whereas, the model predicted a yield of 35 tonnes/ha. The results of the study showed good agreement between predicted and measured corm yields during the four years, which indicated that the calibrated model can be used to improve NPK fertilizer recommendations for elephant foot yam in India. Based on the results and using soil fertility maps and agro ecological unit maps, SSNM zonation maps and secondary and micronutrient fortified customised fertilizers were developed for major elephant foot yam growing environments of India.

Thí nghiệm được tiến hành trong vụ Đông Xuân (ĐX) 2009-2010 và Hè Thu (HT) 2010, sau đó ứng dụng mô hình trong vụ ĐX 2010-2011 và HT 2011. Thí nghiệm được bố trí theo thể thức khối hoàn toàn ngẫu nhiên với 4 nghiệm thức (ô khuyết N bón đầy đủ P, K; ô khuyết P bón đầy đủ N, K; ô khuyết K bón đầy đủ N, P; ô bón đầy đủ N, P, K) và 4 lần lặp lại. Sau khi tìm được công thức phân tiến hành ứng dụng cho mô hình bằng cách chia đôi ruộng nông dân, bón phân theo địa điểm chuyên biệt (SSNM) và QTND. Kết quả thí nghiệm cho biết được lượng dinh dưỡng N, P, K nội tại do đất cung cấp cho một hecta vụ ĐX là 65 kg N + 33 kg P2O5 + 115 kg K2O và công thức phân bón N, P, K đề xuất 90 kg N + 36 kg P2O5 + 22 kg K2O/ha; vụ HT là 49 kg N + 26 kg P2O5 + 88 kg K2O và công thức phân bón N, P, K đề xuất 85 kg N + 40 kg P2O5 + 28 kg K2O/ha. Năng suất lúa ở mô hình bón phân theo SSNM cao hơn cách bón theo QTND là 0,33-0,48 tấn/ha. Trong vụ ĐX 2010-2011 và HT 2011, QTND đã bón lượng phân bón cao hơn so với SSNM, lượng N từ 6 - 9 kg/ha; lượng P2O5 từ 13-18 kg/ha; lượng K2O từ 27-28 kg/ha.

Behaviour of crops and soils due to non-redressal of spatial variability in soil properties is not expressed uniformly within bigger orchard/field which has been the major triggering factor for not able to break yield barrier through conventional nutrient management strategy. Better response of site specific nutrient management (SSNM) over recommended doses of fertilizers (RDF) including those of farmers’ usual fertilizer practices (FFP) signals a wake-up call to address the fertilizer requirements on the principles of SSNM, if the full potential productivity of perennial crops has to be realized and sustain on any given soil type. The SSNM also warrants to tailor the fertilizer application on the basis of spatial variation in soil fertility and tree canopy size within an orchard/field through variable rate fertilizer application to minimize the gap between actual and potential productivity of crops. Application of geospatial tools like GPS and GIS coupled with DRIS- based nutrient diagnostics has substantially aided in developing rationale of fertilizer use using SSNM concept in crops like citrus, avocado, coconut, olive etc. which could well be expanded to other perennial crops.

Fertilizer K and P requirements for rice ( Oryza sativa L.) can be determined with site-specific nutrient management (SSNM) using estimated target yield, nutrient balances, and yield gains from added nutrient. We used the QUEFTS (QUantitative Evaluation of the Fertility of Tropical Soils) model with >8000 plot-level observations to estimate the relationship between grain yield and nutrient accumulation in above-ground dry matter of irrigated rice with harvest index ≥ 0.4. Predicted reciprocal internal efficiencies (RIEs) at 60–70% of yield potential corresponded to plant accumulation of 14.6 kg N, 2.7 kg P, and 15.9 kg K per tonne of grain yield. These RIEs enable determination of plant requirements for K and P and net output of K and P in harvested grain and removed crop residues at a target yield. Yield gains for nutrient applied to irrigated rice averaged 12% for K and 9% for P for 525 to 531 observations. For fields without certain yield gain, fertilizer K and P requirements can be determined by a partial maintenance approach (i.e., fertilizer input < output in nutrient balance), which considers nutrient supply mediated through soil processes and balances trade-offs between financial loss with full maintenance rates and risk of excessive nutrient depletion without nutrient application. When yield gains to an added nutrient are certain, partial maintenance plus yield gain can be used to determine fertilizer requirements. The SSNM-based approach and algorithms enable rapid development of field-specific K and P management. [ABSTRACT FROM AUTHOR]

Wheat is the third most important cereal crop in Nepal after rice and maize both in area and production, but its productivity of 2.3 tonne ha - 1 is very less compared to other developed countries (6 tonne ha -1 for Switzerland and China) in the world. The main cause of low wheat yield in Nepal is the improper and inadequate use of fertilizer devoid of site specific nutrient management practices. Therefore, a farmers' field experiment was conducted during November 2015 to April 2016 to rectify the best fertilizer management options at two sites of Damak and Gauradaha in Jhapa district in eastern-Terai of Nepal using Nutrient Expert®-Wheat model. The research was accomplished in Randomized Complete Block Design with 2 treatments and 20 replications, considering farmers' field as replication. Two treatments included in the experimentation were NE® (Nutrient Expert Recommendation) and FFP (Farmer's Fertilizer Practices). The statistical result revealed the highly significant difference in terms of number of effective tiller m -2 , plant height, filled grain per spike, spike length, grain, straw and biological yields and harvest index. The highest yield (4.71 tonne ha -1 ) was obtained from NE field than FFP (2.99 tonne ha -1 ). On an average, NE based practices produced 58 % higher yield in comparison to FFP. NE based treatment produced significantly higher biomass yield, yield attributes and cost-benefit ratio than FFP treatments. Field experiment validation confirmed that the Nutrient Expert® Wheat model could be used as the most adoptable and practical precision decision support system tool to make a more authentic fertilizer recommendation in eastern-Terai of Nepal., (© 2020 The Author(s).)