Multi-scale traffic vehicle detection based on faster ReCNN with NAS optimization and feature enrichment

It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively difficult to detect vehicles of various scales in traffic scene images,because the vehicles partially obscured by green belts,roadblocks or other vehicles,as well as influence of some low illumination weather.In this paper,we present a model based on Faster ReCNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes.First,we proposed a Retinex-based image adaptive correction algorithm(RIAC)to enhance the traffic images in the dataset to reduce the influence of shadow and illumination,and improve the image quality.Second,in order to improve the feature expression of the backbone network,we conducted Neural Architecture Search(NAS)on the backbone network used for feature extraction of Faster ReCNN to generate the optimal cross-layer connection to extract multi-layer features more effectively.Third,we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets,and improve the robustness of the model for challenging targets such as small scale and severe occlusion.In the implementation of the model,K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model.Our model has been trained and tested on the UN-DETRAC dataset,and the obtained results indicate that our method has art-of-state detection performance.

The Features of Sedimentary Facies and Copper Enrichment Metallogenic Regularities of Kuzigongsu Group in Sareke Glutenite Type Copper Deposits,Wuqia,Xinjiang

1 Introduction Sareke glutenite-type copper deposit is the large size copper deposit discovered in recent years,and it is located Sarekebayi intracontinental pull-apart basin in the western margin of the Tarim basin.Conglomerate of

Spatial distribution characteristics and main controlling factors of germanium in soil of northern Dabie Mountains, China

With the increasing application of germanium(Ge)elements in modern industry,military and medical health industries,especially with the growing demand for Ge-rich agricultural products,the study of Ge-rich soil has become particularly important,but the enrichment pattern and control factors of Ge-rich soil are still not well understood due to the high dispersion and high migration of Ge-rich soil.In this paper,495 surface soil(0-20 cm)and 149 deep soil(150-200 cm)samples were collected from the northern foothills of Dabie Mountain using a double-layer grid layout,and the spatial distribution and enrichment characteristics of Ge were studied by high-resolution method,and the controlling factors affecting the distribution of Ge-rich soil was analyzed by geo-statistics and spatial analogy.The results show an average Ge content of 1.34 mg/kg for the surface and 1.36 mg/kg for the deep soil.In the assessment grade classification of surface and deep soil for Ge,the abundant and sub-abundant grades account for 37.97%and 31.70%,respectively,covering 752 km2 and 634 km2.Surface Ge-rich regions are distributed in concentrated strips in the north-central part of the studied region,and there is no clear pattern in the spatial distribution of deep soils.In the areas under study,such as Fenlukou,Dingji,and Jiangjiadian,the surface soil is very rich in Ge and has a high enrichment factor,which is valuable for agricultural development.In surface soils,river deposits and shallow metamorphic rock parent materials have the highest content of Ge,while in deep soils,the highest content has been found in the parent material of moderately acidic rock.Both surface and deep soils have the highest Ge content in purple paddy soils and plain areas.The source of Ge in the soils of the study area is most influenced by the lithology of the soil-forming parent material,while the distribution of Ge in the surface soils is jointly influenced by pH,SiO_(2),TFe_(2)O_(3),and Al_(2)O_(3) in the soil.This study has implications for understanding the enrichment pattern of Ge in soil and its controlling factors as well as for the development of Ge-rich agricultural products.

Resolving the influence of nitrogen abundances on sediment organic matter in macrophyte-dominated lakes, using fluorescence spectroscopy

A controlled experiment was designed to resolve the influence of nitrogen abundance on sediment organic matters in macrophyte-dominated lakes using fluorescence analysis.Macrophyte biomass showed coincident growth trends with time, but different variation rates with nitrogen treatment. All plant growth indexes with nitrogen addition（N, NH4Cl100, 200, 400 mg/kg, respectively） were lower than those of the control group. Four humiclike components, two autochthonous tryptophan-like components, and one autochthonous tyrosine-like component were identified using the parallel factor analysis model. The results suggested that the relative component changes of fluorescence in the colonized sediments were in direct relation to the change of root biomass with time. In the experiment, the root formation parameters of the plants studied were significantly affected by adding N in sediments, which may be related to the reason that the root growth was affected by N addition.Adding a low concentration of N to sediments can play a part in supplying nutrients to the plants. However, the intensive uptake of NH4^＋may result in an increase in the intracellular concentration of ammonia, which is highly toxic to the plant cells. Hence, our experiment results manifested that organic matter cycling in the macrophyte-dominated sediment was influenced by nitrogen enrichment through influencing vegetation and relevant microbial activity.