A Deep Learning Enabled Multi-Class Plant Disease Detection Model Based on Computer Vision

In this paper, a deep learning enabled object detection model for multi-class plant disease has been proposed based on a state-of-the-art computer vision algorithm. While most existing models are limited to disease detection on a large scale, the current model addresses the accurate detection of fine-grained, multi-scale early disease detection. The proposed model has been improved to optimize for both detection speed and accuracy and applied to multi-class apple plant disease detection in the real environment. The mean average precision (mAP) and F1-score of the detection model reached up to 91.2% and 95.9%, respectively, at a detection rate of 56.9 FPS. The overall detection result demonstrates that the current algorithm significantly outperforms the state-of-the-art detection model with a 9.05% increase in precision and 7.6% increase in F1-score. The proposed model can be employed as an effective and efficient method to detect different apple plant diseases under complex orchard scenarios.

No Evidence of Unexpected Transgenic Insertions in T1190 – A Transgenic Apple Used in Rapid Cycle Breeding – Following Whole Genome Sequencing

Rapid cycle breeding uses transgenic early flowering plants as crossbreed parents to facilitate the shortening of breeding programs for perennial crops with long-lasting juvenility. Rapid cycle breeding in apple was established using the transgenic genotype T1190 expressing the BpMADS4 gene of silver birch. In this study, the genomes of T1190 and its non-transgenic wild-type PinS (F1-offspring of ‘Pinova’ and ‘Idared’) were sequenced by Illumina short-read sequencing in two separate experiments resulting in a mean sequencing depth of 182× for T1190 and 167× for PinS. The sequencing revealed 8,450 reads, which contain sequences of ≥20 bp identical to the plant transformation vector. These reads were assembled into 125 contigs, which were examined to see whether they contained transgenic insertions or if they are not using a five-step procedure. The sequence of one contig represents the known T-DNA insertion on chromosome 4 of T1190. The sequences of the remaining contigs were either equally present in T1190 and PinS, their part with sequence identity to the vector was equally present in apple reference genomes, or they seem to result from endophytic contaminations rather than from additional transgenic insertions. Therefore, we conclude that the transgenic apple plant T1190 contains only one transgenic insertion, located on chromosome 4, and shows no further partial insertions of the transformation vector.Accession Numbers: JQ974028.1.

PERFORMANCE COMPARISON OF UNSUPERVISED SEGMENTATION ALGORITHMS ON RICE, GROUNDNUT, AND APPLE PLANT LEAF IMAGES

This paper focuses on plant leaf image segmentation by considering the aspects of various unsupervised segmentation techniques for automatic plant leaf disease detection. The segmented plant leaves are crucial in the process of automatic disease detection, quantification, and classification of plant diseases. Accurate and efficient assessment of plant diseases is required to avoid economic, social, and ecological losses. This may not be easy to achieve in practice due to multiple factors. It is challenging to segment out the affected area from the images of complex background. Thus, a robust semantic segmentation for automatic recognition and analysis of plant leaf disease detection is extremely demanded in the area of precision agriculture. This breakthrough is expected to lead towards the demand for an accurate and reliable technique for plant leaf segmentation. We propose a hybrid variant that incorporates Graph Cut (GC) and Multi-Level Otsu (MOTSU) in this paper. We compare the segmentation performance implemented on rice, groundnut, and apple plant leaf images for various unsupervised segmentation algorithms. Boundary Displacement error (BDe), Global Consistency error (GCe), Variation of Information (VoI), and Probability Rand index (PRi), are the index metrics used to evaluate the performance of the proposed model. By comparing the outcomes of the simulation, it demonstrates that our proposed technique, Graph Cut based Multi-level Otsu (GCMO), provides better segmentation results as compared to other existing unsupervised algorithms.

The using of gene-oriented SCoT markers in genotyping of the Apple genus (Malus Mill.)

This paper discusses the aspects of optimization of the SCoT genotyping method for representatives of the genus Apple (Malus Mill.), including the orchard apple (Malus domestica). Special attention is paid to the methods of total DNA isolation from apple plant tissue, which is due to the sensitivity of multilocus marker systems to the quality of the nucleic acid preparation. On a sample of total DNA isolated from the leaves of the Golden Delicious cultivar, 18 SCoT markers were tested, from which 4 Malus Mill cultivars and species most promising for genotyping were selected.

Classification of Plant Leaf Diseases using CNN

Agricultural productive is the dominant issue, which affects the economy of the country excessively. So detection of diseases in plants plays a major role in Agricultural field. In previous day’s farmers in the fields used to observe the plants just by seeing with their eye for identification of a disease. But this method may take lot of time, expensive and inaccurate. So advanced technology that can identify plant diseases as easily as possible is needed, in order to decrease the percentage rate of the contamination of crops and increase the fertility. Here in this paper techniques like preprocessing, segmentation and classification of image are used. Here Tomato, Maize, Grape, Potato and Apple plant leaves are used, where different diseases are identified for each plant. For Classification we used Convolution Neural Network Algorithm, so that we can automatically detect the plant leaf diseases. And this will help farmers to identify their diseases as early as possible.

Pengaplikasian Kayu Apu (Pistia stratiotes L) Dalam Menurunkan Kadar BOD, COD dan TSS Pada Limbah Cair Laboratorium Di RSUD Besuki Kabupaten Situbondo

Hospital Laboratory wastewater contains organic compounds with a high enough concentration and possibly contain other chemical compounds and pathogenic microorganisms that can cause disease to the surrounding community. A way to overcome this problem is to do hospital laboratory liquid waste treatment. One alternative that can be used is the phytoremediation method using apu wood plants. Research on decreasing levels of BOD, COD and TSS in Besuki Public Hospital laboratory wastewater using Apu Wood (Pistia stratiotes L) was carried out in January - March 2019 with the aim to determine the decrease in laboratory wastewater BOD, COD and TSS after being given Apu (Pistia stratiotes L). The method used in treating liquid waste laboratory is phytoremediation. The results showed that after being treated as many as 10 treatments with the contact time of the apple plant for 1 day, 3 days and 6 days there was a significant decrease in BOD levels in the treatment with indigo P = 0,000 <0,05 for COD levels based on ANOVA output, known the value based on Anova sig table results is 0,000 <0,05 so it can be concluded that the average COD level after being treated is "DIFFERENT" significantly while the TSS parameter is ANOVA test with sig value 0,309> 0,05 which means "NO DIFFERENCE "Significantly on the treatment. It is recommended for further research, it should be endeavored to use plants that have resistance to waste, especially for contact time so that they can use higher waste concentrations with longer contact times. Keywords: BOD, COD, TSS, Apu Wood Plant (Pistia stratiotes L

SILICON NUTRITION IN ALLEVIATING SALT STRESS IN APPLE PLANT

Salinity is one of the major environmental stresses that adversely affect fruit yield and quality. Thus, finding an effective way of ameliorating salinity damage is important for sustainable fruit production. Silicon (Si) treatment effectively counteracts the effects of many abiotic stress factors such as salinity, drought, cold, iron deficiency. To probe into the potential alleviating salinity malignant effects, we investigated the protective roles of Si. An apple plant (Malus domestica Borkh.) cv. ‘Fuji’ grafted on M9 clonal rootstock was chosen for the experiment and imposed to salinity stress for 4 months with 35 mM NaCl. Si with different three doses (0.5, 1 ve 2 mM) was applied to the roots of the salt-stressed apple plants except control. Si treatments inhibited apple plants growth depression through increasing stomatal conductivity, chlorophyll and decreasing electrolyte leakage. 0.5 mM Si improved root:shoot dry weight under salinity condition. The lowest values of membrane permeability were found in 0.5 and 2 mM Si treatments (21.45 and 21.55%, respectively) while salt had the highest value (48.43%). Salt exhibited a rapid decrease in stomatal conductivity by 49% compared to the control. We hypothesis that Si treatment contributed to cell walls, involving membrane stabilizing and fortification. Our findings showed that Si increased apple plant tolerance against salinity.

Aplikasi Pendeteksi Penyakit Pada Daun Tanaman Apel Dengan Metode Convolutional Neural Network

According to 2017 statistical fruit and vegetable crops published by BPS, total apple production in 2017 amounted to 319004 tons. There are many diseases that can attack apple plants, therefore early detection and identification of plant diseases are the main factors to prevent and reduce the spread of apple plant diseases. CNN method is used in this study with LeNet-5 architecture which can process 3151 imagery data with a mini-mum accuracy level of 75%. This study uses a dataset derived from PlantVillage created by SP Mohanty CEO & Co-founder of CrowdAI with a total of 3151 leaf images that have been classified according to their respec-tive classes. CNN stages include Convolution Layer, Rectified Linear Unit (ReLU), Subsampling, Flattening, Fully Connected Layer. The test results are evaluated using image testing data. The evaluation process is done using a confusion matrix. Based on the results of testing applications that are designed with 99,4% model ac-curacy and 97,8% validation accuracy, the application is useful for detecting apple disease using apple leaf images.