Nondestructive determination of soluble solids and firmness in mix-cultivar melon using near-infrared CCD spectroscopy

Nondestructive evaluation of melon quality is in great need of comprehensive study.Soluble solids content(SSC)and firmness are the two indicators of melon internal quality that mostly a®ect consumer acceptance.To provide guidance for fruit classification,internal quality standards was preliminarily established through sensory test,as:Melon with SSC over 12
Brix,firmness 4–5.5 kgf·cm^(-2)2 were considered as satisfactory class sample;and SSC over 10
Brix,¯rmness 3.5–6.5 kgf·cm^(-2) as average class sample.The near infrared(NIR)nondestructive detection program was set as spectra collected from the stylar-end,Brix expressed by the average SSC of inner and outer mesocarp,each cultivar of melon was detected with its own optimum integration time,and the second derivative algorithm was used to equalize them.Using wavelength selected by genetic algorithms(GA),a robust SSC model of mix-cultivar melon was established,the root mean standard error of cross-validation(RMSECV)was 0.99 and the ratio performance deviation(RPD)nearly reached 3.0,which almost could meet the accuracy requirement of 1.5
Brix.Firmness model of mix-cultivar melon was acceptable but inferior.

A novel nondestructive detection approach for seed cotton lint percentage using deep learning

Background The lint percentage of seed cotton is one of the most important parameters for evaluating seed cotton quality and affects its price.The traditional measuring method of lint percentage is labor-intensive and time-consuming;thus,an efficient and accurate measurement method is needed.In recent years,classification-based deep learning and computer vision have shown promise in solving various classification tasks.Results In this study,we propose a new approach for detecting the lint percentage using MobileNetV2 and transfer learning.The model is deployed on a lint percentage detection instrument,which can rapidly and accurately determine the lint percentage of seed cotton.We evaluated the performance of the proposed approach using a dataset comprising 66924 seed cotton images from different regions of China.The results of the experiments showed that the model with transfer learning achieved an average classification accuracy of 98.43%,with an average precision of 94.97%,an average recall of 95.26%,and an average F1-score of 95.20%.Furthermore,the proposed classification model achieved an average accuracy of 97.22%in calculating the lint percentage,showing no significant difference from the performance of experts(independent-sample t-test,t=0.019,P=0.860).Conclusion This study demonstrated the effectiveness of the MobileNetV2 model and transfer learning in calculating the lint percentage of seed cotton.The proposed approach is a promising alternative to traditional methods,providing a rapid and accurate solution for the industry.

Determination of potassium sorbate and sorbic acid in agricultural products using THz time-domain spectroscopy

t The aim of this study was to investigate the feasibility of detecting potassium sorbate(PS)and sorbic acid(SA)in agricultural products using THz time-domain spectroscopy(THz-TDS).The absorption spectra of PS and SA were measured from 0.2 to 1.6 THz at room temperature.The main characteristic absorption peaks of PS and SA in polyethylene and powdered agricultural products with different weight ratios were detected and analyzed.Interval partial least squares(iPLS)combined with a particle swarm optimization and support vector classification(PSO-SVC)algorithm was proposed in this paper.iPLS was used for frequency optimization,and the PSO-SVC algorithm was used for spectrum analysis of the preservative content based on the optimal spectrum ranges.Optimized PSO-SVC models were obtained when the THz spectrum from the PS/SA mixture was divided into 11 or 12 subintervals.The optimal penalty parameter c and kernel parameter g were found to be 1.284 and 0.863 for PS(0.551-1.487 THz),1.374 and 0.906 for SA(0.454-1.216 THz),respectively.The preliminary results indicate that THz-TDS can be an effective nondestructive analytical tool used for the quantitative detection of additives in agricultural products.

Nondestructive perception of potato quality in actual online production based on cross-modal technology

Nowadays,China stands as the global leader in terms of potato planting area and total potato production.The rapid and nondestructive detection of the potato quality before processing is of great significance in promoting rural revitalization and augmenting farmers’income.However,existing potato quality sorting methods are primarily confined to theoretical research,and the market lacks an integrated intelligent detection system.Therefore,there is an urgent need for a post-harvest potato detection method adapted to the actual production needs.The study proposes a potato quality sorting method based on cross-modal technology.First,an industrial camera obtains image information for external quality detection.A model using the YOLOv5s algorithm to detect external green-skinned,germinated,rot and mechanical damage defects.VIS/NIR spectroscopy is used to obtain spectral information for internal quality detection.A convolutional neural network(CNN)algorithm is used to detect internal blackheart disease defects.The mean average precision(mAP)of the external detection model is 0.892 when intersection of union(IoU)=0.5.The accuracy of the internal detection model is 98.2%.The real-time dynamic defect detection rate for the final online detection system is 91.3%,and the average detection time is 350 ms per potato.In contrast to samples collected in an ideal laboratory setting for analysis,the dynamic detection results of this study are more applicable based on a real-time online working environment.It also provides a valuable reference for the subsequent online quality testing of similar agricultural products.

Nondestructive detection of infertile hatching eggs based on spectral and imaging information

In order to quickly distinguish infertile eggs from fertile eggs,the hyperspectral imaging technology consisting of imaging and spectral information was used for detecting the fertile information of eggs.Before hatching eggs were incubated,a hyperspectral imaging system(wavelength between 400 to 1000 nm)was used to acquire the images one-by-one manually.The characteristic information of ratios of length to short axis,elongation,roundness and the ratios of the yolk area to the whole area was extracted based on the images.The normalization method was used as the spectral data preprocessing,and then 155 spectral characteristic variables were extracted from 520 nm waveband through the correlation coefficient method.Principal component analysis(PCA)method was adopted to reduce the dimensions of image-spectrum fusion information;the top six principal components were extracted.Support vector machine(SVM)method was used to establish classification of fertile and infertile eggs models,which are based on image,spectrum and image-spectrum fusion information respectively.The accuracy rates of the SVM models were 84.00%,90.00%and 93.00%respectively.The experimental results show that the model based on image-spectrum fusion information technology is superior to the single information model.Hyperspectral transmission imaging technology is effective and feasible to detect the fertile hatching eggs before incubation.

Detection of rice seed vigor by low-field nuclear magnetic resonance

A new method to predict the seed vigor of rice was developed to control adulteration during the seed trading process and to address the deficiencies of traditional manual detection methods.Low-field nuclear magnetic resonance(LF-NMR)technique was used to detect the vigor of rice seeds.Four varieties(Beijing-1,Qianchonglang-2,Yanfeng-47 and Shennong-265)of rice seeds from the Rice Research Institute of Shenyang Agricultural University were chosen for the experiment.The transverse relaxation time T_(2),T_(21) and T_(22) were observed in the experiment.The peak start time of free water(transverse relaxation time T_(22)),signal amplitude of bound water(transverse relaxation time T_(21)),and moisture content decreased with the decrease in the vigor of the seeds.There were no obvious trends observed for the top of the peak and the end point of the transverse relaxation time T_(22).In addition,the start,top,and end time of the peak(transverse relaxation time T_(21)),and the signal amplitude of bound water showed no consistent changes.The results indicated that LF-NMR could be used as a method to distinguish the vigor of rice seeds rapidly.This study provided theoretical basis and technical support for the rapid detection of rice seed vigor.

Online automatic grading of salted eggs based on machine vision

The quality of salted eggs differs in curing process.They need to be tested and graded before factory packaging.The dynamic images of salted eggs were acquired on conveyor.Firstly,preprocessing of color images must be done:the target area of the binary image was determined by mathematical morphology and removal of the object of a small area.According to the binary image is a convex or concave figure,the target region light leaked or not was determined.The effects of leaked region were eliminated by searching for mutation points,fitting salted egg boundary by the Least Square algorithm,labeling the binary image and extracting single target area.Then,six characteristic parameters were extracted in color space,and quality testing model was established by minimum error probability.The experimental results indicated that the detection accuracy reached above 93%and classification efficiency was 5400/h.It is proved the model is feasible for salted egg grading.

Early diagnosis and monitoring of nitrogen nutrition stress in tomato leaves using electrical impedance spectroscopy

Nitrogen(N)is a life element for crop growth.In tomato growth and development,N stress often occurs and degrades crop yield and quality.Superfluous N can noticeably increase the nitrate content,which can be degraded into strong carcinogenic substance-nitrite.An accurate and timely monitoring and diagnosis of nutrition during crop growth is premise to realize a precise nutrient management.Crop N monitoring methods have been developed to improve N fertilizer management,and most of them are based on leaf or canopy optical property measurements.Although many optical/spectral plant N sensors have already commercialized for production use,low accuracy for phosphorus(P)and potassium(K)detection and diagnosis remains an important drawback of these methods.To explore the potential of N diagnosis by electrical impedance and perform study for nutrition status of plant NPK meanwhile by the electrical impedance,it is necessary that evaluate the N nutrition level by leaf impedance spectroscopy.Electrical impedance was applied to determine the physiological and nutritional status of plant tissues,but few studies related to plant N contents have been reported.The objective of this study was to evaluate the N nutrition level by leaf impedance spectroscopy and realize the early diagnosis and monitoring of N nutrition stress in tomato.Five sets of tomato plant samples with different N contents were cultivated in a Venlo greenhouse.N content of leaves was determined,and electrical impedance data were recorded in a frequency range of 1 Hz to 1 MHz.The obtained impedance data were analyzed using an equivalent circuit model for cellular tissues.The variation of equivalent parameters along with N content was analyzed,and the sensitive impedance spectroscopy characteristics of N nutrition level were extracted.Furthermore,the effect of moisture content on impedance measurement was discussed and the prediction model for N content was developed.Results showed that electrical impedance can be conveniently applied to early diagnosis and monitoring for tomato N nutrition stress.

Discrimination of brownheart of Korla pear using vibration frequency spectrum technique

The purpose of this work was to use a nondestructive method for detecting the brownheart of Korla pear to reduce the chance of infection among pears without brownheart.A mechanical impulse method based on vibration testing system was used to excite the fruits.The consistent acquisition signal indicated that the test is repeatable at the same positions of fruit equator(cheek).A remarkable frequency signal was excited at 9-12 N force by a rubber tipped hammer.The dominant frequency was identified at the maximum response magnitude to assess the internal defect,and the result was not influenced by the distances between the defect borders and the excitation points.The sharp increase of defect mass could significantly affect the dominant frequency.Relationship between the dominant response frequency(fd)and the defect mass percentage(ω)was characterized by an equation fd=410.649e-0.0833ω+261.947 with a good correlation coefficient(R2=0.925).A defect mass of 2.281%was determined as a discrimination threshold.Once the threshold exceeded 2.281%,the defective pear could be classified with a high accuracy rate of 96.7%.This finding would provide guidance for determining the optimal detecting time to the brownheart of Korla pears,according to the specific storage conditions when the vibration frequency spectrum method is deployed.

Study of imaging unknown objects by cosmic-ray muons

Introduction Cosmic-ray muon imaging is a kind of nondestructive detection technology which can be used to detect unknown objects in geological exploration,civil engineering and nuclear safety.Transmission imaging and scattering tomography schemes are studied.Method The transmission scheme uses a multilayer detector to measure the direction of a cosmic-ray muon passing through an object.The scattering scheme involves placing two detectors upstream and downstream of the object to record the incident and exit directions of the muon passing through the object.The effect of the detector resolution on the imaging clarity of transmission imaging was studied.The applicable scenarios of the two schemes were analyzed.Results The results by calculating show that in the transmission imaging of a hundred-meter object,a spatial resolution of 2.5 m can be achieved,and Cu and Fe can be discriminated with a density difference of 1.1 g/cm3.Scattering tomography is mainly suitable for meter-level objects,which can detect 0.2 m chamber and distinguish 0.05 m heavy metal blocks in rock.

Predicting bruise susceptibility in apples using Vis/SWNIR technique combined with ensemble learning

Bruise susceptibility in fruits is an important indicator in evaluating risk factors for bruising caused by external factors.Prediction of the bruising susceptibility of fruit can provide useful information for proper postharvest handling and storage operations.In this study,visible and shortwave near-infrared(Vis/SWNIR)technique was used to develop nondestructive method for predicting the bruise susceptibility of apples.Vis/SWNIR spectra covering 400-1100 nm were collected for 300‘Golden Delicious’apples over a time period of three weeks after harvest.A pendulum-like device was used to simulate impact bruise at three impact energy levels of 1.11 J,0.66 J and 0.33 J.Bruise volumes were estimated from the digital images of the bruised apples by using the bruise thickness model.Three prediction models,i.e.partial least squares model(PLS),partial least squares model combined with successful projection algorithm(SPA-PLS),and selective ensemble learning based on feature selection(SELFS),for bruise susceptibility were developed for each impact energy level as well as for the pooled data.Compared with PLS and SPA-PLS model,SELFS gave the better prediction results for bruise susceptibility,with the correlation coefficient of R_(p)=0.800-0.886 for the prediction set,the root-mean-square error of 38.7-62.1 mm^(3)/J for the prediction set(RMSEP),and the residual predictive deviation(RPD)of 1.78-2.14 for three impact energy level.For three impact energy levels,the RMSEP and RPD value obtained by SELFS model improved by 14.8%-20.0%and 15.0%-24.5%compared to PLS model,and 11.4%-21.2%and 11.5%-27.1%compared to SPA-PLS model,respectively.The SELFS model achieved relatively lower prediction accuracies for the pooled data,with the R_(p) values of 0.731,RMSEP of 85.46 mm^(3)/J,and RPD of 1.46,which were also better than that of PLS model and SPA-PLS model.This research demonstrated that Vis/SWNIR technique combined with ensemble learning is promising technique for rapid assessment of bruise susceptibility of fruit,which would be useful for postharvest handling of fruit.