A Systematic Literature Review of Machine Learning and Deep Learning Approaches for Spectral Image Classification in Agricultural Applications Using Aerial Photography

Recently,there has been a notable surge of interest in scientific research regarding spectral images.The potential of these images to revolutionize the digital photography industry,like aerial photography through Unmanned Aerial Vehicles(UAVs),has captured considerable attention.One encouraging aspect is their combination with machine learning and deep learning algorithms,which have demonstrated remarkable outcomes in image classification.As a result of this powerful amalgamation,the adoption of spectral images has experienced exponential growth across various domains,with agriculture being one of the prominent beneficiaries.This paper presents an extensive survey encompassing multispectral and hyperspectral images,focusing on their applications for classification challenges in diverse agricultural areas,including plants,grains,fruits,and vegetables.By meticulously examining primary studies,we delve into the specific agricultural domains where multispectral and hyperspectral images have found practical use.Additionally,our attention is directed towards utilizing machine learning techniques for effectively classifying hyperspectral images within the agricultural context.The findings of our investigation reveal that deep learning and support vector machines have emerged as widely employed methods for hyperspectral image classification in agriculture.Nevertheless,we also shed light on the various issues and limitations of working with spectral images.This comprehensive analysis aims to provide valuable insights into the current state of spectral imaging in agriculture and its potential for future advancements.

Warhead fragments motion trajectories tracking and spatio-temporal distribution reconstruction method based on high-speed stereo photography

High speed photography technique is potentially the most effective way to measure the motion parameter of warhead fragment benefiting from its advantages of high accuracy,high resolution and high efficiency.However,it faces challenge in dense objects tracking and 3D trajectories reconstruction due to the characteristics of small size and dense distribution of fragment swarm.To address these challenges,this work presents a warhead fragments motion trajectories tracking and spatio-temporal distribution reconstruction method based on high-speed stereo photography.Firstly,background difference algorithm is utilized to extract the center and area of each fragment in the image sequence.Subsequently,a multi-object tracking(MOT)algorithm using Kalman filtering and Hungarian optimal assignment is developed to realize real-time and robust trajectories tracking of fragment swarm.To reconstruct 3D motion trajectories,a global stereo trajectories matching strategy is presented,which takes advantages of epipolar constraint and continuity constraint to correctly retrieve stereo correspondence followed by 3D trajectories refinement using polynomial fitting.Finally,the simulation and experimental results demonstrate that the proposed method can accurately track the motion trajectories and reconstruct the spatio-temporal distribution of 1.0×10^(3)fragments in a field of view(FOV)of 3.2 m×2.5 m,and the accuracy of the velocity estimation can achieve 98.6%.

Multi-class classification of pathological myopia based on fundus photography

Pathological myopia(PM)is a severe ocular disease leading to blindness.As a traditional noninvasive diagnostic method,fundus color photography(FCP)is widely used in detecting PM due to its highfidelity and precision.However,manual examination of fundus photographs for PM is time-consuming and prone to high error rates.Existing automated detection technologies have yet to study the detailed classification in diagnosing different stages of PM lesions.In this paper,we proposed an intelligent system which utilized Resnet101 technology to multi-categorically diagnose PM by classifying FCPs with different stages of lesions.The system subdivided different stages of PM into eight subcategories,aiming to enhance the precision and e±ciency of the diagnostic process.It achieved an average accuracy rate of 98.86%in detection of PM,with an area under the curve(AUC)of 98.96%.For the eight subcategories of PM,the detection accuracy reached 99.63%,with an AUC of 99.98%.Compared with other widely used multi-class models such as VGG16,Vision Transformer(VIT),E±cientNet,this system demonstrates higher accuracy and AUC.This artificial intelligence system is designed to be easily integrated into existing clinical diagnostic tools,providing an e±cient solution for large-scale PM screening.

Fundus photography,fundus fluorescein angiography,and optical coherence tomography of healthy cynomolgus monkey,New Zealand rabbit,Sprague Dawley rat,and BALB/c mouse retinas

Background:A variety of experimental animal models are used in basic ophthalmological research to elucidate physiological mechanisms of vision and disease pathogenesis.The choice of animal model is based on the measurability of specific parameters or structures,the applicability of clinical measurement technologies,and the similarity to human eye function.Studies of eye pathology usually compare optical parameters between a healthy and altered state,so accurate baseline assessments are critical,but few reports have comprehensively examined the normal anatomical structures and physiological functions in these models.Methods:Three cynomolgus monkeys,six New Zealand rabbits,ten Sprague Dawley(SD)rats,and BALB/c mice were examined by fundus photography(FP),fundus fluorescein angiography(FFA),and optical coherence tomography(OCT).Results:Most retinal structures of cynomolgus monkey were anatomically similar to the corresponding human structures as revealed by FP,FFA,and OCT.New Zealand rabbits have large eyeballs,but they have large optic disc and myelinated retinal nerve fibers in their retinas,and the growth pattern of retinal vessels were also different to the human retinas.Unlike monkeys and rabbits,the retinal vessels of SD rats and BALB/c mice were widely distributed and clear.The OCT performance of them were similar with human beings except the macular.Conclusions:Monkey is a good model to study changes in retinal structure associated with fundus disease,rabbits are not suitable for studies on retinal vessel diseases and optic nerve diseases,and rats and mice are good models for retinal vascular diseases.These measures will help guide the choice of model and measurement technology and reduce the number of experimental animals required.

Fundus photography,fluorescein angiography,optical coherence tomography and electroretinography of preclinical animal models of ocular diseases

The eye is an immune-privileged and sensory organ in humans and animals.Anatomical,physiological,and pathobiological features share significant similarities across divergent species(1).Each compartment of the eye has a unique structure and function.The anterior and posterior compartments of the eye contain endothelium(cornea),epithelium(cornea,ciliary body,iris),muscle(ciliary body),vitreous and neuronal(retina)tissues,which make the eye suitable to evaluate efficacy and safety of tissue specific drugs(2).

A New Chromatic Vision:The Early Impact of Color Photography on the Representation of Architecture

The role of color photography in the representation of architecture is a subject little investigated by architectural historiography.The link between the color values of architectural design and its visual transmission in the early phase of modernism was certainly problematic.Color photography had an undeniable impact on architectural color in practice:color photographs in books and periodicals published between the 1940s and 1960s clearly influenced the use of color in architectural design.

Investigating Jet Appearance in Different Background Lights with High-Speed Frame Photography

Shaped charge jet formation process is studied under the conditions of different background lights by means of high speed frame photography. In order to shoot true jet appearance, the glass tube in which jet moves is vacuumized. The experiment results show that observing jet appearance with the double reflecting mirrors system is feasible as long as the vacuum of the glass tube can meet the requirement of experiment.

THE APPLICATION OF FOURIER TRANSFORM IN REGULARIZED SPECKLE PHOTOGRAPHY

Two speckle patterns are recorded using a video cameracorresponding to undisplaced and displaced states into the memory ofa computer. After one pattern is added to another, the displacementof the measured points can be calculated by using the Fouriertransform. The expression of displacement field is deduced fornon-uniform displacement, and as an example, a rotatanle disc istested for illustration. The re- sults prove that the present methodis promising.

Problems and improvement strategies of photography education in Colleges and Universities

In recent years, with the development of social economy, college photography education career has been developing rapidly, but at the same time, some problems have appeared. In the construction of photography specialty, a series of problems, such as unreasonable course arrangement, outdated hardware facilities and imperfect teaching software system, are hindering the de-velopment of photography education in universities. Therefore, college educators should seriously sum up the problems existing in college photography education, and carry out photography teaching reasonably and scientifically.

On the Improvement Method of Photography Practice Teaching

Photography teaching is different from general culture teaching.There are obvious differences in teaching methods and teachers' teaching styles.To this end,photography teaching should be innovative,really interested in students,love the photography course,and put photography knowledge into practice.At present,some photography teachers have some problems in the way of explai-ning photography knowledge,which makes it difficult for students to really grasp the practical knowledge of photography,and is not conducive to arousing students' interest in learning. The reason is that photography teachers do not attach importance to practical teaching.

The research and develop on the training mode of the innovative professional photography personnel

Under the background of the Digital Age, This article proposes suggestions on the training mode of the innovative professional photography personnel. We should rest on the good foundations of the demand pattern of ＇Double art talent＇ and take advantage of the advanced special services of Art Design. We should persist in the principle of Artistic Innovation first. We should establish an optimal training scheme of ＇photography creative talents＇ . A highly innovative photography subject system will train talents for ＇Design of photography＇ . We should change our opinions on traditional teaching, and to suggest improvements or comment on the practicability of the innovation and reform of Professional photography.

A Model for Estimating Total Forest Coverage with Ground-Based Digital Photography

Vegetation fractional coverage (VFC) is one of the key indicators of vegetation distribution. In the work a measurement-based model was developed to derive total forest VFC (TG) as well as the VFC of trees (T) and shrub-grasses (G) separately in a subtropical forest area in Nanjing, China. Both upward and downward photographs were taken with a digital camera in 72 quadrats (10 m × 10 m each). Fifteen models were established and validated. Models jointly using both T and G performed better than those using the T and G separately. The best model, TG = T + G- 1.134 × T × G- 0.025 (R2 = 0.9115, P < 0.01, root mean squared error = 0.0789), is recommended for application. This model provides a good way to obtain total forest VFC values through taking tree and shrub-grass photos on ground below tree canopy rather than above tree canopy.

Full-field mapping of internal strain distribution in red sandstone specimen under compression using digital volumetric speckle photography and X-ray computed tomography

It is always desirable to know the interior deformation pattern when a rock is subjected to mechanicalload. Few experimental techniques exist that can represent full-field three-dimensional （3D） straindistribution inside a rock specimen. And yet it is crucial that this information is available for fully understandingthe failure mechanism of rocks or other geomaterials. In this study, by using the newlydeveloped digital volumetric speckle photography （DVSP） technique in conjunction with X-ray computedtomography （CT） and taking advantage of natural 3D speckles formed inside the rock due to materialimpurities and voids, we can probe the interior of a rock to map its deformation pattern under load andshed light on its failure mechanism. We apply this technique to the analysis of a red sandstone specimenunder increasing uniaxial compressive load applied incrementally. The full-field 3D displacement fieldsare obtained in the specimen as a function of the load, from which both the volumetric and the deviatoricstrain fields are calculated. Strain localization zones which lead to the eventual failure of the rock areidentified. The results indicate that both shear and tension are contributing factors to the failuremechanism. 2015 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting byElsevier B.V. All rights reserved.

Does bird photography affect nest predation and feeding frequency?

Bird photography is a popular and growing form of ecotourism that contributes to the economic growth of local communities,but its disturbance to bird reproduction remains understudied.We worked in a tropical forest of southern China,which has experienced a sharp increase in the number of photographers in recent years.We compared nests that were photographed and those that were not,in their nest predation and parental feeding rates.Including nests of 42 species,the results demonstrate that the predation rate of nests that were not photographed(incubation stage:43.3%of 194 nests;nestling stage:34.5%of 110 nests)was significantly higher than that of photographed ones(incubation:2.4%of 83 nests;nestling:11.1%of 81 nests).Among different nest types,open cup nests in shrub and trees were most affected by photography,in both incubation and nestling stages.Of five factors investigated,including three natural factors(nest stage,structure and position),and two anthropogenic factors(photography and distance to forest edge),only photographic disturbance and nest structure had significant effects(open nests had higher predation).The feeding frequency at nests when photographers were present was not strongly different from when they were not present.Human activity therefore had no negative effects on the birds,but showed a positive effect on their nesting success,in terms of reducing nest predation rates.However,there needs to be further assessment of other aspects of nesting(e.g.,clutch size,duration of nestlings in nests),and other kinds of stress responses(e.g.,hormonal changes),before the total effect of bird photography can be understood.

Computer-Assisted DEM-Based Aero Photography Design

In aerial photography, the primary factor is terrain undulation. However, most of the external aerial photography software used for aerial photography design do not take terrain undulation influence into consideration. Therefore, the design result has comparative randomicity and ＂gaps＂ are expected. An aerial photography design system is developed by analyzing the terrain undulation influence to the design result with DEM data so that the forward overlap and side overlap can be justified according to the block terrain undulation to meet specifications or standards. The data designed by this system is compared with the real flying data. The results show that making use of DEM to assist in aerial photography design can ensure that the designed result fits the real terrain better.

Analysis of droplet transfer of pulsed MIG welding based on electrical signal and high-speed photography

In order to study how welding parameters affect welding quality and droplet transfer, a synchronous acquisition and analysis system is established to acquire and analyze electrical signal and instantaneous images of droplet transfer simultaneously, which is based on a self-developed soft-switching inverter. On the one hand, welding current and voltage signals are acquired and analyzed by a self-developed dynamic wavelet analyzer. On the other hand, images are filtered and optimized after they are captured by high-speed camera. The results show that instantaneous waveforms and statistical data of electrical signal contribute to make an overall assessment of welding quality, and that optimized high-speed images allow a visual and clear observation of droplet transfer process. The analysis of both waveforms and images leads to a further research on droplet transfer mechanism and provides a basis for precise control of droplet transfer.

Study on Gardening Photography

Through analyzing the concept,development history and works of gardening photography,unique status and charm of gardening photography in the field of photography art was explored.It was proposed that professional knowledge about gardening disciplines should be grasped,and the shoot targets shouid be fully investigated to present excellent works of gardening photography.In addition,photographers should have strong artistic sense,and be able to grasp the means of expressing image language,so that they will produce works with outstanding themes,proper light,concise composition,harmonious hue color and shadow tone.Sophisticated equipment and post processing skills are also needed.

Evaluating the impact of sampling schemes on leaf area index measurements from digital hemispherical photography in Larix principis-rupprechtii forest plots

Background:Digital hemispherical photography(DHP)is widely used to estimate the leaf area index(LAI)of forest plots due to its advantages of high efficiency and low cost.A crucial step in the LAI estimation of forest plots via DHP is choosing a sampling scheme.However,various sampling schemes involving DHP have been used for the LAI estimation of forest plots.To date,the impact of sampling schemes on LAI estimation from DHP has not been comprehensively investigated.Methods:In this study,13 commonly used sampling schemes which belong to five sampling types(i.e.dispersed,square,cross,transect and circle)were adopted in the LAI estimation of five Larix principis-rupprechtii plots(25m×25 m).An additional sampling scheme(with a sample size of 89)was generated on the basis of all the sample points of the 13 sampling schemes.Three typical inversion models and four canopy element clumping index(Ωe)algorithms were involved in the LAI estimation.The impacts of the sampling schemes on four variables,including gap fraction,Ωe,effective plant area index(PAIe)and LAI estimation from DHP were analysed.The LAI estimates obtained with different sampling schemes were then compared with those obtained from litter collection measurements.Results:Large differences were observed for all four variable estimates(i.e.gap fraction,Ωe,PAIe and LAI)under different sampling schemes.The differences in impact of sampling schemes on LAI estimation were not obvious for the three inversion models,if the fourΩe algorithms,except for the traditional gap-size analysis algorithm were adopted in the estimation.The accuracy of LAI estimation was not always improved with an increase in sample size.Moreover,results indicated that with the appropriate inversion model,Ωe algorithm and sampling scheme,the maximum estimation error of DHP-estimated LAI at elementary sampling unit can be less than 20%,which is required by the global climate observing system,except in forest plots with extremely large LAI values(~>6.0).However,obtaining an LAI from DHP with an estimation error lower than 5%is impossible regardless of which combination of inversion model,Ωe algorithm and sampling scheme is used.Conclusion:The LAI estimation of L.principis-rupprechtii forests from DHP was largely affected by the sampling schemes adopted in the estimation.Thus,the sampling scheme should be seriously considered in the LAI estimation.One square and two transect sampling schemes(with sample sizes ranging from 3 to 9)were recommended to be used to estimate the LAI of L.principis-rupprechtii forests with the smallest mean relative error(MRE).By contrast,three cross and one dispersed sampling schemes were identified to provide LAI estimates with relatively large MREs.

Robust Watermarking of Screen-Photography Based on JND

With the popularity of smartphones,it is often easy to maliciously leak important information by taking pictures of the phone.Robust watermarking that can resist screen photography can achieve the protection of information.Since the screen photo process can cause some irreversible distortion,the currently available screen photo watermarks do not consider the image content well and the visual quality is not very high.Therefore,this paper proposes a new screen-photography robust watermark.In terms of embedding region selection,the intensity-based Scale-invariant feature transform(SIFT)algorithm used for the construction of feature regions based on the density of feature points,which can make it more focused on the key content of the image;in terms of embedding strength,the Just noticeable difference(JND)model is applied to limit the intensity of the watermark embedding according to the luminance and texture of the picture to balance robustness and invisibility;after embedding watermark,the coefficients in the neighborhood are again adjusted with optimal constraints to improve the accuracy of watermark extraction.After experiments,it is shown that the method we proposed can improve the correct rate of watermark extraction,the quality of the visual aspect of the watermarked picture is also improved.

The Semiotics of Photography： Towards Objective Hermeneutics

How can we know what a photograph means？ What messages or codes exist semiotically that help us to interpret it？ The mass consumption of images is reflected in the large number of contests, scholarships, galleries, museographic exhibitions and especially in advertising： What are the semiotic parameters that can be established for a reliable and clear judgment to discriminate one photo from another？ For critics or judges, artists or the general public, this article seeks to answer these questions by proposing a semiotic analysis based on new advances in science, in psychology of perception and neurobiology of vision; said methodology, currently being applied intensively at ESAV and used in the postgraduate studies of the present author, seeks to facilitate the understanding of the meaning--to improve the hermeneutics---ofthe photographic image in a systematic and objective way.