Tomato detection method using domain adaptive learning for dense planting environments

This study aimed to address the challenge of accurately and reliably detecting tomatoes in dense planting environments,a critical prerequisite for the automation implementation of robotic harvesting.However,the heavy reliance on extensive manually annotated datasets for training deep learning models still poses significant limitations to their application in real-world agricultural production environments.To overcome these limitations,we employed domain adaptive learning approach combined with the YOLOv5 model to develop a novel tomato detection model called as TDA-YOLO(tomato detection domain adaptation).We designated the normal illumination scenes in dense planting environments as the source domain and utilized various other illumination scenes as the target domain.To construct bridge mechanism between source and target domains,neural preset for color style transfer is introduced to generate a pseudo-dataset,which served to deal with domain discrepancy.Furthermore,this study combines the semi-supervised learning method to enable the model to extract domain-invariant features more fully,and uses knowledge distillation to improve the model's ability to adapt to the target domain.Additionally,for purpose of promoting inference speed and low computational demand,the lightweight FasterNet network was integrated into the YOLOv5's C3 module,creating a modified C3_Faster module.The experimental results demonstrated that the proposed TDA-YOLO model significantly outperformed original YOLOv5s model,achieving a mAP(mean average precision)of 96.80%for tomato detection across diverse scenarios in dense planting environments,increasing by 7.19 percentage points;Compared with the latest YOLOv8 and YOLOv9,it is also 2.17 and 1.19 percentage points higher,respectively.The model's average detection time per image was an impressive 15 milliseconds,with a FLOPs(floating point operations per second)count of 13.8 G.After acceleration processing,the detection accuracy of the TDA-YOLO model on the Jetson Xavier NX development board is 90.95%,the mAP value is 91.35%,and the detection time of each image is 21 ms,which can still meet the requirements of real-time detection of tomatoes in dense planting environment.The experimental results show that the proposed TDA-YOLO model can accurately and quickly detect tomatoes in dense planting environment,and at the same time avoid the use of a large number of annotated data,which provides technical support for the development of automatic harvesting systems for tomatoes and other fruits.

幂合变换与幂合矩阵

把满足A3=A的矩阵A叫做幂合矩阵,满足A3=A的线性变换A叫做幂合变换。显然,幂和矩阵(变换)是幂等矩阵(变换)和是对合矩阵(变换)的统一和推广。讨论了它们的性质,并给出了它们的等价条件。

Variation of Water-Use Efficiency of Leymus chinensis and Cleistogenes squarrosa in Different Plant Communities in Xilin River Basin, Nei Mongol

Water is usually considered to be a key limiting factor for the growth and reproduction of steppe plants in the Xilin River Basin, Nei Mongol. Foliar delta C-13 values, an indicator of long-term intercellular carbon dioxide concentration and thus of water-use efficiency (WUE) in plants, were measured on Leymus chinensis (Trin.) Tzvel. and Cleistogenes squarrosa (Trin.) Keng. in six communities of different habitats in die Xilin River Basin. The foliar delta C-13 values of both species tended to increase with decreasing soil water content (SWC) and a significant negative correlation was found between foliar delta C-13 Values and SWC in different soil layers, indicating that the two species could change WUE according to water availability. We also found relatively constant leaf water contents (LWC) of the two species in different habitats. Our results implied that the two steppe species might have adapted to different soil water regimes either through adjusting stomatal conductance to get a proper WUE, or through enhancing the osmosis-regulating ability to keep a relatively stable LWC. Our findings could partially explain why the two plant species have a wide distribution range and become dominant in the Xilin River Basin.

Effects of Different Vegetable Planting Modes on Soil Microbial Flora and Enzyme Activity

To investigate the effects of different vegetable growing regions and planting modes on soil quality,soils in high,medium and low altitude areas of Guizhou were respectively sampled under different vegetable efficient planting modes,and the variations of soil microbial flora and enzyme activities were analyzed. The soil microbial count and total bacteria of the vegetable efficient cultivation mode were significantly higher than that of the control （traditional planting mode） in each planting area,and the microbial diversity index was also improved to varying de- grees.The soil phosphatase,catalase and urease activities of the vegetable efficient planting mode were higher than that of the control.The soil catalase and urease activities were higher than that of the control by 1.37-1.44 and 1.51-2.80 times. Application of vegetable efficient planting mode in different regions will help to im- prove the soil quality in a given period.

Investigation and Risk Assessment of Alien Invasive Plants in Riparian Zone of Dongjiang River

[Objective] The aim was to conduct investigation and risk assessment of the alien invasive plants in riparian zone of Dongjiang River.[Method] The field survey was carried out by the combination methods of sampling plots investigation and route survey,and the whole Dongjiang River was covered.52 sampling points were set in the main streams,and 87 sampling points in tributaries streams,624 quadrats were set in total;multi-index comprehensive evaluation method was adopted for the risk assessment with 6 first-level indexes and 25 second-level indexes.[Result] Through the investigation on the riparian zone of Dongjiang River for 2 years,51 species of alien invasive plants were found,belonging to 17 families and 38 genera.The risk level assessment results of these 51 species showed that 18 species,such as Ageratum conyzoides,belonged to high-level risk with over 60 points;26 species,such as Chenopodium ambrosioides,were between 30 and 60 points which belonged to middle-level risk;7 species,such as Paspalum dilatatum,with less than 30 points belonged to low-level risk.[Conclusion] Basic data on the protection,water environmental management and functional zoning of riparian zone of Dongjiang River were provided.

Vegetation Evolution in the Northern South China Sea Region Since 40 ka BP - An Attempt to Reconstruct Palaeovegetation Based on Biomization

Vegetation evolution in the northern South China Sea region since 40 500 a BP is reconstructed using biomization procedure based on pollen data from deep sea core 17940. The result shows that, it is feasible to reconstruct palaeovegetation using biomization procedure, when pollen, particularly Pinus pollen, transported by mind over long distance is excluded. Results from factor analysis suggest that humidity and temperature are the two main factors determining vegetation evolution on land around the northern South China Sea. From 40 500 a BP to 11 100 a BP, broad-leaved evergreen forest (WAMF), and montane conifers(MGRF) occurred on hills and low mountains; while steppe (STEP) predominated on the exposed shelf. The main feature of the vegetation evolution is the frequent alternation between MGRF (or WAMF) and STEP, implying abrupt changes in millennium scale between humid/cold and dry/temperate climate. All abrupt climate events could be broadly correlated with Henrich events and Dansggard-Oscherge events. One of the events around 12 700 a BP, sees the occurrence of MGRF, suggesting that climate turned humid and cold rapidly. This may be correlated with the Younger Dryas event; Broad-leaved evergreen (WAMF) predominates since 11 000 a BP. During the early Holocene and late Holocene tropical rainforest (TRFO) or tropical seasonal forest (TSFO) occurred several times.

Effects of Vegetation Coverage and Management Practice on Soil Nitrogen Loss by Erosion in a Hilly Region of the Loess Plateau in China

Soil erosion and nutrient loss due to erosion are world-wide problems. Similar to soil loss by erosion, soil nitrogen (N) loss by erosion in small catchments is affected by vegetation coverage. The practice of comprehensive management for catchments mainly by adjusting cropland, grassland and woodland areas was widely adopted to reduce soil and water loss in catchments of the Chinese Loess Plateau. Three experiments under natural and artificial rainfall conditions on N loss by erosion for a model catchment and for an actual catchment in Zhifanggou of Ansai County in China was performed to determine the relationships between comprehensive management and N loss by runoff in small catchments. The results for vegetation coverage of 60%, 40%, 20% and 0 show that runoff loss of ammonium, nitrate, and total N were 87.08, 44.31, 25.16, 13.71 kg/km(2); 85.50, 74.06, 63.95, 56.23 kg/km(2); and 0.18, 1.18, 1.98, 7.51 t/ km(2), respectively. Due to reduction in the size of cropped area on steeply sloping land, soil N loss by erosion in the catchments was decreased by 15.8% as compared with that in 1992, i.e., from 8 758.5 kg in 1992 to 7 562.2 kg in 1998. Whereas, catchments act as a filter for ammonium and nitrate in rain, the catchment filtering effects on nitrate is remarkably higher than that on ammonium. The enrichment of < 20 mum aggregate in sediment results in the enrichment of organic matter and total N in flood sediment. Greater vegetation coverage can effectively decrease soil erosion and total N loss. However, soil mineral N loss increased as vegetation coverage increased.

Nitrate Nitrogen Content Characteristics in Groundwater of Typical Planting Areas in Liao River Basin

The contents of nitrate nitrogen（NO-3-N） in underground water from typical planting areas in Liao river basin were analyzed, so as to provide theoretical basis for rational fertilization and effectively prevent the NO-3-N content from exceeding standard. The results showed that difference of the contents of NO-3-N in groundwater from different typical planting areas was significant. The highest content of NO-3-N in underground water was 37.4 mg/L from flower-growing region, then 22. 3 mg/L from maize-growing region, 21.9 mg/L from vegetable-growing region, and the lowest content of NO-3-N in underground water was 19.2 mg/L from rice-growing region. Except rice-growing region, the contents of NO-3-N in underground water of the samples in all planting areas were exceeding standard limit; potential health risk still existed in rice-growing region. Accordingly 12.5%-87.5%, 9.4%-75.5%, 17.9%-58.9% and 21.4%-96.0% of the samples were exceeding standard limit in maize growing region, rice-growing region, vegetable-growing region and flower-growing region. The contents of NO-3-N in under-groundwater before the rainy season was higher than that of NO-3-N in under-groundwater after the rainy season at the same depth of the well.

Phytoplankton in Prydz Bay and Its Adjacent Sea Area of Antarctica During the Austral Summer (1998/1999)

The cell density, species composition and distribution of phytoplankton, and their relations to environmental factors in Prydz Bay and its adjacent sea area, Antarctica (69degrees-77degrees E, 62degrees-70degrees S) during the austral summer of 1998/1999 were investigated. A total of 48 taxa belonging to 21 genera of phytoplankton in the sea area were identified. The average cell density of phytoplankton was 22.46 x 10(3) cells/dm(3), of which diatoms were predominant (84.51%). The highest cell density of phytoplankton occurred in Prydz Bay and the adjacent continental shelf where the average cell density was 46.03 x 10(3) cells/dm(3). The lowest cell density (3.34 x 10(3) cells/dm(3)) occurred in deep sea area. The dominant species of phytoplankton was Fragilariopsis curta. The vertical distribution of phytoplankton density was highest in the upper part of 0-50 in depth, lower in 100 in and lowest in 150 in. The species composition and cell density of phytoplankton were influenced by water circulation. The cell density was positively correlated with water temperature and salinity, and negatively correlated with the concentration of nutrients.

Investigation on Physiological Status of Regional Vegetation Using Pushbroom Hyperspectral Imager Data

To extract vegetation pigment concentration and physiological status has been studied in two test areas covered with swamp and flourish vegetation using pushbroom hyperspectral imager (PHI) data which flied in September of 2000 at Daxing'anling district of Heilongjiang Province, China. The ratio analysis of reflectance spectra (RARS) indices, which were put forward by Chappelle et al (1992), are chosen in this paper owing to their effect and simpleness against both comparison with various methods and techniques for exploration of pigment concentration and characteristics of PHI data. The correlation coefficients between RARS indices and pigment concentration of vegetation were up to 0.8. The new RARS indices modes are established in the two test areas using both PHI data and spectra of different vegetations measured in the field. The indices' parameter images of chlorophyll a (Chl a), chlorophyll b (Chl b) and carotenoids (Cars) of the test areas covered with swamp and flourish vegetation are acquired by the new RARS indices modes. Furthermore, the regional concentration of Chl a and Chl b are extracted and quantified using regression equations between RARS indices and pigment concentrations, which were built by Blackburn (1998). The results showed the physiological status and variety clearly, and are in good agreement with the distribution of vegetation in the field.

Selection of Indigenous Plants in Roads in Northwestern Loess Plateau Area

[Objective] The aim was to study on resource utilization of indigenous plants in roads in order to solve resource shortage of indigenous plant in bioengi- neering. [Method] Based on northwestern loess plateau area, we conducted research on selection of indigenous plants and survey on the species which could live in adverse condition, mainly by surveying, with frequent tests as a supplement, to make a comprehensive observation and evaluation on plant combinations which are to be extensively applied. [Result] The research indicated that in central dividing strips, four arbors, including Platycladus orientalis （Linn.）Franco, Sabina chinensis, Sabinachinen- sis（Linn.）Ant., and Ulmus purnila L., and nine frutices, including Lycium chinense, Tamarix austromongolica, Rosa xanthina Lindl., Forsythia suspense, Atriplex canescens, Amygdalus triloba, Hippophae rharnnoides Linn., Arnorpha fruticosa L., and Syzygium aromaticum could be planted. In addition, Gleditsia sinensis Lain, Robinia pseudoacacia L, Elaeagnus angustifolia L., Rhus typhina, Salix babylonica, Salixmatsudana and Fraxinus chinensis, tall and strong, are suitable to be planted as border trees, presenting a good vision. Frutices, limited by height, are usually with more branches, which will provide a good landscape if grown with tall plants. Prunus dacidiana and Ulmus pumila L., prone to be damaged by insects, should be planted with other species to stop the infestation in space. [Conclusion] The research laid foundation for selection of indigenous plants in roads, providing references for resource shortage of indigenous plant for bioengineedng.

Comparison of Soil Respiration in Two Grass-Dominated Communitiesin the Xil in River Basin: Correlations and Controls

A comparative study of soil respiration was conducted between in a semi-arid steppe community and in a wet meadow community in the Xilin River Basin of Nei Mongol. The seasonal pattern, the climatic controls, and the correlations of soil respiration with plant biomass components, were examined for each community. The main results are reported as follows: (1) The seasonal changes in soil respiration in the two communities had similar dynamic patterns (both being of two peaks), ranging from 312.8 to 1738.9 mg C(.)m(-2).s(-1) and from 354.6 to 2235.6 Mg C.m(-2).s(-1) in the growing season for the steppe plot and the meadow plot respectively. The soil respiration rate of the meadow plot was distinctly higher than that of the steppe plot, with the daily averages being 1349.6 mg C-m(-2).s(-1) and 785.9 mg C-m(-2).s(-1) respectively. (2) The correlation between soil respiration rate and soil moisture was much more significant than with temperature for the steppe community, and being on the contrary for the meadow community, reflecting the different effects of the two climatic factors in different habitats. Based on these regressive relations, the total CO2 efflux rate in the growing season in 2001 was estimated as 142.4 g C/m(2) in the steppe plot, and 236.1 g C/m(2) in the meadow plot. (3) There was no evident relation between the total canopy biomass and CO2 evolution rate, but a significant power function relation between the live canopy biomass and CO2 evolution rate in the meadow plot was detected. In the steppe plot, there existed only a weak relation between soil respiration and either live or total canopy biomass.

Effects of Land Use Change on Ecosystem Services in Arid Area Ecological Migration

Ecological migration is the process of increasing the population density in the immigration area and transferring the ecological pressure from emigration area to immigration area. This process may result in significant changes in land use and land cover in the area of immigration and have an important effect on ecosystem services. Therefore, scientifically revealing the effects and differentiation mechanisms of ecological migration on ecosystem services is becoming an important issue related to the implementation of the national ecological migration strategy in China. This study employed the Hongsibu District as a typical example of ecological migration. Hongsibu District is located in the central Ningxia steppe and desert steppe areas. Remote sensing data covering five periods from the period before ecological migration in 1995 and after migration in 2000, 2005, 2010, and 2015 was used to measure the value of ecosystem services(ESV). A geographical detector model and the value of ecosystem services model were used to diagnose the dynamic mechanism of the effects of land use change on ecosystem services. The results showed that: 1) The development of large-scale ecological resettlement has caused the area of cultivated land and urbanized land area to increase significantly in the area of immigration, while the grass area decreased significantly. 2) The overall value of the Hongsibu ecosystem services increased in a form of a ‘V'. Among them, during the period of 1995–2005, the overall ESV decreased and had an annual rate of change of-0.67%. During the period of development 2005–2015, the ESV increased steadily, with an annual rate of change of 0.79%. 3) The proportion and total ESV in soil formation and protection, waste treatment, and biodiversity conservation of the Hongsibu District decreased from 57.61% in 1995 to 56.17% in 2015, indicating that the region's ecological regulation function slightly decreased. 4) The ESV in the Hongsibu District, showed a low distribution pattern of ecosystem services increasing from northeast to southwest, and the capacity of three townships, Hongsibu, Taiyangshan, and Liuquan, to provide ecosystem services gradually declined over time. The ecological service function of Xinzhuangji Township and Dahe Township gradually improved. 5) The sensitivity index of the ESV of each land use type was less than 1, indicating that the environment lacks flexibility in providing a strong ESV index in Hongsibu, which shows that the research results are reliable and believable. 6) During the study period, the decisive force of the change of land use on ecosystem services in Hongsibu District was: grassland(0.9934), climate regulation(0.9413), soil formation and protection(0.9321) and waste treatment(0.9241).

Least squares reverse-time migration in the pseudodepth domain and reservoir exploration

Least squares reverse-time migration （LSRTM） is an inversion method that removes artificial images and preserves the amplitude of reflectivity sections. LSRTM has been used in reservoir exploration and processing of 4D seismic data. LSRTM is, however, a computationally costly and memory-intensive method. In this study, LSRTM in the pseudodepth domain was combined with the conjugate gradient method to reduce the computational cost while maintaining precision. The velocity field in the depth domain was transformed to the velocity field in the pseudodepth domain; thus, the total number of vertical sampling points was reduced and oversampling was avoided. Synthetic and field data were used to validate the proposed method. LSRTM in the pseudodepth domain in conjunction with the conjugate gradient method shows potential in treating field data.

Plant Developmental Biology in China: Past, Present and Future

Plant development is a process from germination of seed to vegetative growth, flower initiation and development, fertilization and finally to the next generation seed formation. A lot of achievements have been obtained in plant developmental biology in China. Since the 1990's, those young generation scientists returned to China from abroad to use molecular and genetic techniques to study morphological, physiological and biochemical process of plant development. The present paper reviews the progress in some research area of plant developmental biology in the past decades and also prospects the chance and future of plant developmental studies, due to the recent advances of plant genome sequencing and functional genomics in China under the international research background.

Evaluation of Spectral Scale Effects in Estimation of Vegetation Leaf Area Index Using Spectral Indices Methods

Spectral index methodology has been widely used in Leaf Area Index(LAI) retrieval at different spatial scales. There are differences in the spectral response of different remote sensors and thus spectral scale effect generated during the use of spectral indices to retrieve LAI. In this study, PROSPECT, leaf optical properties model and Scattering by Arbitrarily Inclined Layers(SAIL) model, were used to simulate canopy spectral reflectance with a bandwidth of 5 nm and a Gaussian spectral response function was employed to simulate the spectral data at six bandwidths ranging from 10 to 35 nm. Additionally, for bandwidths from 5 to 35 nm, the correlation between the spectral index and LAI, and the sensitivities of the spectral index to changes in LAI and bandwidth were analyzed. Finally, the reflectance data at six bandwidths ranging from 40 to 65 nm were used to verify the spectral scale effect generated during the use of the spectral index to retrieve LAI. Results indicate that Vegetation Index of the Universal Pattern Decomposition(VIUPD) had the highest accuracy during LAI retrieval. Followed by Normalized Difference Vegetation Index(NDVI), Modified Simple Ratio Indices(MSRI) and Triangle Vegetation Index(TVI), although the coefficient of determination R^2 was higher than 0.96, the retrieved LAI values were less than the actual value and thus lacked validity. Other spectral indices were significantly affected by the spectral scale effect with poor retrieval results. In this study, VIUPD, which exhibited a relatively good correlation and sensitivity to LAI, was less affected by the spectral scale effect and had a relatively good retrieval capability. This conclusion supports a purported feature independent of the sensor of this model and also confirms the great potential of VIUPD for retrieval of physicochemical parameters of vegetation using multi-source remote sensing data.

Relationship between High-quality Tea and the Geological Background in Qionglai Mountains

[Objective] Many strata emerge in the high-quality tea producing area in Qionglai Mountains,which belongs to Mesozoic Cretaceous(K),Mingshan Fm(E1-2mn)and Dayi conglomerate Team(N)of Cenozoic Tertiary System as well as Quaternary System(Q).Analysis on the migrating features of soil heavy metal from each stratum to tea leaves showed that the enrichment ability of elements was Cu>Cd>Zn>Ni>Hg>As>Cr>Pb.By comparing the soil background values with the soil base values of Sichuan hilly area and Chengdu economic region,it could be concluded that Cd,Hg,Mo,Pb and Se were enriched,and other metal elements existed at a level close to that of the comparing areas.The pH was a little acidic,and high quality tea cultivation could be developed here at present.With the increase of human activities,the environment of study area tends to deteriorate.Therefore,land planning and use should be reasonable.The government should control the environment nearby and prevent pollution.

Spatio-temporal Variability of Soil Water at Three Seasonal Floodplain Sites: A Case Study in Tarim Basin,Northwest China

The floodplain -egetation of the Tarim River in Northwest China is strongly influenced by irrigated agriculture. The abstrac- tion of river water disturbs; the natural dynamics of the floodplain ecosystem. The human impact on the hydrological system by bank dams and the irrigation of cotton plantings have caused adverse changes of the Tarim River and its floodplains, so the current stocks of the typical Tugai vegetation show significant signs of degradation. Field studies of soils and statistical analysis of soil moisture data have shown that the vitality of the Tugai vegetation is primarily determined by its position to the riverbank and the groundwater. There exist complex interactions between soil hydrological conditions and the vitality of the vegetation. But the availability of water is not only influenced by the groundwater level and seasonal flood events. The spatial distribution of stocks at different states of vitality seems also to be decisively influenced by physical soil properties. Our results show that the water supply of plant communities is strongly in- fluenced by the soil texture. Spatial differences of soil moisture and corresponding soil water tensions may be the decisive factors for the zonafion of vegetation. Physical soil properties control the water retention and rising of capillary water from deeper soil layers and the phreatic zone and may supply the root systems of the phreatophytic vegetation with water. Keywords： soil moisture;soil texture; soil water tensions; Tarim River; water retention

Study on Optical Properties of Unpigmented Suspended Particles, Yellow Substance and Phytoplankton Algae in Taihu Lake

The optical properties of a waterbody are subjects of research on limnological hydrooptics. The optical properties of, and parameterization methods for, three impurities in Taihu Lake, namely unpigmented suspensions, yellow substance and phytoplankton algae, are analyzed in detail in this paper dealing with the optical types of Taihu Lake waters both in winter and in summer. The results showed that: 1) The optical parameters of suspended particles can be calculated by the application of Mie Theory; 2) The absorption of yellow substance decreases exponentially with increasing wavelength, with the mean value of decreasing rate being about 0.014 nm -1; 3) Taihu Lake waters appeared to be of type PY in winter, with suspended particles and yellow substance being the dominant light absorbers, and type CPY in summer, with suspended particles, yellow substance and phytoplankton algae being the dominant light absorbers; 4) In winter or summer, unpigmented suspension has the main effect on the scattering property of Taihu Lake waters.

Seasonal Variability of Phytoplankton Community Structure in Relation to Different Nitrogen-Phosphorus Ratios in the Southern Coastal Waters of Zhejiang, China

With the rapid development of economy and increase of population in the drainage areas, the nutrient loading has increased dramatically in the Changjiang estuary and adjacent coastal waters. To properly assess the impact of nutrient enrichment on phytoplankton community, seasonal microcosm experiments were conducted during August 2010-July 2011 in the coastal waters of Zhejiang Province. The results of the present study indicated that the chl a concentration, cell abundance, diversity indices, species composition and community succession of the phytoplankton varied significantly with different N/P ratios and seasons. Higher growth was observed in the 64：1 （spring）, 32：1 （summer）, 16：1 （autumn） and 128：1,256：1 （winter） treatments, respectively. The values of Shannon-Wiener index （H） and Pielou evenness index （J） were lower in the 8：1 and 16：1 treatments in autumn test, while H value was higher in the 128：1 and 8：1 treatments in winter test. A definite community succession order from diatoms to dinoflagel lares was observed in the autumn and winter tests, while the diatoms dominated the community throughout the culture in the spring and summer tests.