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]

This article presents the findings of a study conducted in West Bengal, India, on the correlation between different tillage practices, nutrient management, and the yield of maize. The study found that there was no significant difference in grain and stover yield between different tillage practices. However, significantly higher grain and stover yield was observed with site-specific nutrient management. The study also revealed positive correlations between various growth parameters and grain yield. These findings provide valuable insights into sustainable farming practices for maize cultivation, contributing to food security and environmental stewardship. [Extracted from the article]

Site-specific nutrient management (SSNM) for a rice (Oryza sativa L.)-growing domain is typically developed by first using the nutrient omission plot technique (NOPT) to determine yield responses with added nitrogen (N), with added phosphorus (P), and with added potassium (K) and then conducting an evaluation experiment to verify an SSNM practice developed from the NOPT results. We hypothesized that an effective SSNM practice could be more rapidly developed through concurrent NOPT experiments and evaluation of a preliminary best-bet SSNM practice derived from experiences in other rice-growing domains. Our study domain was rainfed lowland rice in an Indonesian tidal swampland without past NOPT experiments and without a verified SSNM practice. Concurrent NOPT and evaluation experiments were conducted in the wet season (WS) and dry season across 20–25 farmers' fields either directly or not directly influenced by tides. Rice yields and gross return above fertilizer cost for a preliminary SSNM practice were higher than for farmers' fertilizer practice in both seasons and for two fertilizer practices based on soil tests in WS. The inherent ability of SSNM to adjust fertilizer rates for an attainable target yield and time fertilizer applications to match crop demand contributed to the superiority of SSNM. In NOPT experiments, relative yields without added N, without added P, and without added K respectively averaged 0.70, 0.89, and 0.88 across seasons and tidal locations. These relative yields enabled preliminary SSNM fertilizer rates to be further adjusted for the rice-growing domain. Rice growers in domains without a verified SSNM practice can benefit, before NOPT experiments are completed, by using generic SSNM guidelines for timing fertilizer applications and adjusting fertilizer rates for a target yield. After completion of NOPT experiments, the fertilizer rates can be further adjusted based on measured yield responses with added N, P, or K. [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.

Indiscriminate use of inorganic fertilizers has bought threat to soil health in respect to physical, chemical and biological properties of soil. Therefore, it is necessary to minimize the usage of inorganic fertilizers by substituting with organics. It is well established fact that the improvement in quality and productivity of the crops could be made possible with combined application of organic sources and balanced chemical fertilizers. Therefore, a field experiment was conducted during rabi 2021-22 at College of Agriculture, UAS, Raichur to study the effect of organic sources and nutrient levels on growth, yield and economics of safflower (Carthamus tinctorius L.). Eighteen treatment combinations were evaluated which consisted of three organic sources and six nutrient levels. The experiment was laid out in Factorial RBD design with three replications. The results indicated that among organic sources, application of FYM @ 5 t /ha recorded significantly higher growth parameters viz., plant height (85.95 cm), total dry matter accumulation (121.59 g/plant), seed yield (1467 kg/ha) and stalk yield (2216 kg /ha). Further this treatment recorded higher gross returns (Rs. 80,699/ha) and net returns (Rs. 51247/ha) compared to other treatment. Among the different nutrient levels, SSNM (STCR equation) + micronutrients Zn @ 25 kg/ha + S @ 10 kg/ha showed significant effect on growth parameters and seed yield (1508 kg/ha) as compared to other treatment. Moreover, the combination registered higher gross returns (Rs. 82,964/ha) and net returns (Rs. 54,540/ha) compared to other nutrient levels. [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.

The QUEFTS (QUantitative Evaluation of the Fertility of Tropical Soils) model is a valuable tool for developing location-specific management of macronutrients such as N, P, and K to attain higher yields. A review analysis is conducted to understand the applicability of the QUEFTS model in developing site-specific nutrient management (SSNM) for different crops. A total of 33 published articles are available for the QUEFTS model for different crops across the world. The SSNM is available for rice, wheat, maize, potato, tropical tuber crops (cassava, sweet potato, elephant foot yam, taro, and yam), peanut, watermelon, soybean, radish, tea, sweet potato, elephant foot yam, yam, and taro in different countries using the QUEFTS model, and it is discussed in this review. The maximum values of accumulation and dilution of N, P, and K; the balanced nutrient requirements to produce 1000 kg of crop yield; and their internal efficiencies are discussed in this review paper. The review shows that the QUEFTS model parameters derived for rice, wheat, and maize can be directly adapted to different environmental conditions. The data available for potato demand further refinement of the model for its general application irrespective of the agro-climatic conditions. The two sets of data available for cassava are similar and can use the same model parameters for further studies for nutrient management. Single set of data for other crops shows the necessity for more studies to increase their production. The modeled values for different crops presented in this review can further assist the decision-makers in scheduling fertilizer applications for the respective crops, reducing fertilizer and economic losses, and maintaining soil quality. Thus, QUEFTS can be widely used as a simple model to recommend nutrient management practices. However, the QUEFTS model considers only the mean meteorological parameters irrespective of their daily variations. It also considers the soil as a homogeneous medium, which influences the crop yield estimation, which needs to be addressed among agricultural researchers. Therefore, further modifications are also needed for QUEFTS to incorporate different agro-climatic and management conditions and different interactive processes, including other micronutrients, irrigation practices, and sub-soil properties as this model assumes that nutrients are the only limiting factors in crop yield. This review also demands the necessity to focus on the nutrient management of other food crops to ensure food security. [ABSTRACT FROM AUTHOR]