PowerPoint Presentation

[Audio] Hi guys, today I will show my presentation about AUTOMATED PLANT HEALTH DIAGNOSIS REVOLUTIONIZING DISEASE IDENTIFICATION THROUGH IMAGE PROCESSING TECHNIQUES.

[Audio] The project group members include Siti Nadzirah Mohamed Radzwan (Leader), Nurul Azma Zakaria, Zaheera Zainal Abidin, and Muhammad Zaid Kasbudi. Let's go we discuss our topic today.

[Audio] This is our table of contents for this topic we have introduction, problem statement, contribution/impact, implementation and last commercialization potentail.

[Audio] For introduction, This project aims To use sophisticated image processing techniques for precise automated identification of plant diseases using extensive picture databases. 2. Utilize deep learning and image processing to improve disease diagnosis accuracy. 3. To increase crop yield and ensure the sustainability of hydroponic agriculture in urban areas. 4. To address the need for efficient techniques to identify leaf diseases in hydroponic plants.

[Audio] For problem statement side we have 4 problem we facing for this study is Hydroponic farming lacks effective, scalable, and accurate methods for detecting leaf diseases, Inefficient disease detection can negatively affect plant health and crop yields, Develop solutions using advanced image processing techniques and Enhance accuracy in disease detection and improve monitoring and management practices..

[Audio] For benefit to this study, get Reduce human error in disease identification, Implement preventive measures effectively Optimize plant health and crop yield in hydroponic systems..

[Audio] On project overview Data Collection The data for plant leaf disease detection can be sourced from repositories like Plant Village and Kaggle. These repositories contain a wide range of images of plant leaves with various diseases. Data Preprocessing To Prepare data for training by cleaning , resizing, and augmenting images to increase diversity without collecting new images. Building the Model Convolutional neural networks (CNNs) are used to extract features from image, followed compiling and training the model using preprocessed datasets to minimize loss..

[Audio] Next we get Evaluating the Model Validation involves using a separate validation set to assess model performance during training, while testing evaluates performance on an unseen test dataset to estimate real-world effectiveness. Saving the Model To store the trained model for future use, enabling quick deployment and reuse without the need for retraining. Accuracy Result Measure the accuracy of the model in correctly identifying diseased and healthy leaves. The accuracy result provides a quantitative measure of the model's performance.

[Audio] This is our flow chart Process Start: Identification of input parameters. Retrieval of relevant data from database. Checking data availability. Error notification generation if necessary data is unavailable. Application of transformation logic for data processing. Validation of transformed data. Handling of invalid data. Generation of intermediate output based on validated data. Validation of intermediate output. Logging of validation errors. Finalization of output based on validated intermediate results. End of process..

[Audio] On CONTRIBUTION/IMPACT The proposed automated plant health diagnosis system, utilizing advanced image processing techniques, has the potential to revolutionize agricultural practices by improving disease detection accuracy, yield, and sustainability. This technology benefits farmers and contributes to global food security by ensuring healthier crops and reducing agricultural losses..

[Audio] In this study We follow Sustainable Development Goals based on the United Nations on Climate Action and Life on Land.

[Audio] IMPLEMENTATION we use MATLAB MATLAB is a powerful tool for image processing and machine learning, enabling the development of efficient automated plant health diagnosis, agricultural productivity and sustainability. Kaggle (Plant Village) Dataset Overview Kaggle is a global data hub for data scientists and machine learning practitioners. It hosts thousands of datasets across various industries and offers insights into trends and techniques organized into train, test, and validation folders. CONVOLUTION ARCHITECTURE Convolution and pooling layers are used to extract features from an input image or video, reducing computational efficiency by down sampling the output..

[Audio] For COMMERCIALIZATION POTENTIAL Implementing an automated plant health diagnosis system using advanced image processing techniques for leaf disease detection can significantly improve efficiency, accuracy, and early disease detection, leading to optimised resource use and cost savings. This technology enhances crop management and sustainability, promoting healthier ecosystems and increased agricultural productivity..

[Audio] For RELEVANT AUTHORITIES on this study is Lembaga Kemajuan Pertanian Muda (MADA), Lembaga Pertubuhan Peladang (LPP) , Jabatan Pertanian Malaysia (DOA), Institut Penyelidikan Dan Kemajuan (MARDI), Institut Latihan Pengembangan Pertanian (ILPP), and Lembaga Pemasaran Pertanian Persekutuan (FAMA).

[image]. [image]. A MAGIC PICTURE.

[Audio] "Thank you all for your time and attention today. I hope the insights shared have been valuable and have sparked new ideas or reinforced existing ones. As we move forward, I encourage you to consider how these concepts can be applied to your work and daily life. Once again, thank you for being here today. Have a great day!".