final phase ppt on detection of bilva patra leaf.pptx

Ppt

1. VISVESVARAYA TECHNOLOGICAL UNIVERSITY, BELAGAVI “Vision Based Identification of Bilva Patra leaves and Quality Grading Using Images” Presented By Chandan H S (4YG19CS003) Chandu L N(4YG19CS004) Karthik H C(4YG19CS006) Rohith K(4YG19CS012) Internal Guide Dr. Mahanthesh C Elemmi Ph.D. Assoc. Prof., Dept. of CSE NCE, Hassan. DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING NAVKIS COLLEGE OF ENGINEERING, HASSAN 2022-23 Project Presentation on

2. Contents Introduction Literature Survey. Requirment Specification Implementation Code Result Conclusion

3. Introduction Computer vision is a field of artificial intelligence (AI) that enables computers and systems to derive meaningful information from digital images, videos and other visual inputs and take actions or make recommendations based on that information. The concept of computer vision is based on teaching computers to process an image at a pixel level and understanding it. Healthcare is amongst the one of the basic need required by an individual to stay fit in this world. All health systems share the common goal of improving the health of their population. Mainly, there are two types of treatments Allopathic and Naturopatathy.

4. Cont... Allopathic medicine is also called allopathy. It is a health system in which medical doctors, nurses, pharmacists, and other healthcare professionals are licensed to practice and treat symptoms and diseases. It contains more side effect. Naturopathy is an effective alternative to antibiotics in infectious diseases, producing no toxic side effects and bringing about rapid recovery, Naturopathy medicines are much less expensive than conventional prescription drugs. We have plenty of medicines available from various plants and leaves. in that bilva patra leaves are one of the significant and useful plant.

5. Cont... Bilva patra leaves is used for more than fifty types of significant diseases such as Respiratory problems, Diabetes, Diarrhea & Cholera, Anti-inflammatory, Antimicrobial functions, Constipation, Scurvy . Identification of bilva patra leaves and its nutrition gradient becomes more important to make people to know about its benefits, so that people can use naturally available plants and take care of their health without spending more money.

6. MOTIVATION AND PROBLEM STATEMENT Since, bilva patra leaf is the medicine for more than 50 major diseases there is a need for development of methodology that can automatically identify bilva patra leaf, stages of bilva patra leaves and also the dried conditions of bilva patra leaves. Hence, the work on Vision Based Identification of bilva patra Tree from Leaf Images and Quality Grading is proposed.

7. OBJECTIVES To Classify Leaf image as bilva patra or other.  Bilva patra leaf image into different stages as initial, intermediate or later.  Suggest for Bilva patra leaf based on conditions such as leaves dried under sunlight and under shadow

8. Tree structure of work carried out Fig :Tree-structure showing the different stages of bilva patra leaves.

9. IMAGES OF BILVA PATRA LEAFAND OTHER LEAF Bilva Patra Leaf Other Leaf

10. STAGES OF BILVA PATRA LEAF Initial Stage Intermediate Satge Final Stage

11. DRIED LEAF IMAGES Dried under sunlight Dried under shows

12. LITERATURE SURVEY IMAGE ACQUISITION Image acquisition is the first step of the fundamental steps of DIP. In this stage, an image is given in the digital form. The healthy and affected leaf images are captured by using digital devices like digital cameras and other digital imaging devices. As an initial step of image processing technique, the captured images were uploaded to the system to detect defected leaf [1]. Image acquisition can be widely described as the activity of restoring a picture from some origin, usually a hardware-based source, which can be used to identify various location using various images acquired by digital camera [2]. In [3] Image acquisition by common digital camera to capture pictures of different tomato plants which can be used to detect healthy and unhealthy tomato plant. To quantify various eye movements like lateral, upper and lower eye movements can be done by Image Acquisition from different angle using digital camera [13]. Image is captured from a digital camera though the image might be able to identify different kinds of infected leaf or fruit [14]. In [17] the images acquisition work was done by digital camera, The images were acquired for doing survey of different cotton field like diseased or not.

13. Cont... IMAGE PREPROCESSING The leaves are recognized based on pre-processing during the image processing phase, an Contrast Stretching technique is used to classification of various leaves and quality of leaves image [1]. The different pre-processing technique in image processing are used in cherry leaf image is to analyze and extract the S and V channels of HSV color plane, in S channel image, the leaf can be seen clearly than other channel images and in V channel, diseased part can be seen clearly using pre-processing technique [16]. Preprocessing operations such as Fourier filtering, edge detection and morphological operations are applied to canny leaf images to classify images for further process [12]. There are a variety of image preprocessing technique such as image conversion, image clipping, image enhancement technique and Region of interest technique are used to improve the quality of image in signature detection [15]. The Gradient operator is used for preprocessing to suppress noise or small fluctuation in the image in leaf disease detection [16]. Various preprocessing techniques used such as image cropping, resizing, color transformation, contrast enhancement and filtering is done for removing noise and enhancing images in dataset for detection of various tomato diseases [17]

14. Cont... FEATURE EXTRACTION Various types of feature extraction are possible such as color, texture, shapes, and edges. In there proposed system, they deal with the color and shape of the leaf to get better accuracy[1]. The process of identifying the diseases in plants, features such as color, texture, morphological and color coherence vector are commonly used [7]. Feature extraction for soil erosion due to water and wind, construction hazards and obstacles, and potential open space areas resulted in four sensitivity image [10]. The type of plant leaves and unhealthy region of plant leaves or fruits are determined in feature extraction phase is done by ROI operations [15]. The features are extracted using morphological, texture and color feature extraction method. Thirteen important features are extracted and normalized for Analysis of Neofabraea Leaf Spot in Olive Plant [16]. Features are also extracted using Gray Level Co-occurrence Matrix(GLCM) for diseases identification. GLCM methodology is an second order statistical features extraction system for reviewing surface that considers the spatial relationship of pixels dark level dissemination [17]. Gabor filter method for feature extraction of color, shape,texture are used in feature extraction of leaves [18]. The color space feature extraction respectively is used to reduce effect of illumination and distinguish between chili and non- chili leaf color efficiently [14] .

15. Cont... TRAINING OF THE MODEL Classification and detection techniques that can be used for plant leaf disease classification. Here pre-processing is done before feature extraction. RGB images are converted into grey level image. Then basic Morphological functions are applied on the image to extract the image of vein from each leaf. Then the image is converted into binary image. After that if binary pixel value is 0 it is converted to corresponding RGB image value. Finally by using Pearson correlation and dominating feature set and Naive Bayesian classifier disease is detected.[1] The SSD approach is based on a feed-forward convolutional network that produces a fixed-size collection of bounding boxes and scores for the presence of object class instances in those boxes, followed by a non-maximum suppression step to produce the final detections. In the approach in [2] this network is able to deal with objects of various sizes by combining predictions from multiple feature maps with different resolutions.

16. Cont... CLASSIFICATION TECHNIQUES The classification was performed by analyzing several algorithms in the literature review. SVM and CNN were chosen as the best and appropriate classifier with 93.54% and 93.72% of accuracy detection of plant disease detection respectively [1]. Classification techniques for classifying plant diseases are artificial neural networks (ANN) and support vector machine (SVM) [7]. Supervised classifications were performed in image processing software on a Landsat to identify accuracy rate of soil erosion [10]. The implementation of the hybrid GAACO algorithm can segment not only color leaf image but also fruit image accurately and detect the diseases present in leaf or fruit image can be done using TSVM disease classification as per [15].

17. REQUIREMENTS ANALYSIS HARDWARE REQUIREMENTS: Processor : Intel-Core i5,i7 and above Processor Speed : 2.6 GHz, GPU System RAM : 16 GB Storage : 512 GB to 1TB

18. Cont... SOFTWARE REQUIREMENTS Operating System : Windows 10 Language : Python Tool kit : Jupiter Notebook Front end : HTML, CSS, BOOTSTRAP, JAVASCRIPT Frame work : Django.

19. IMPLEMENTATION Proposed methodology

20. Cont... IMAGE ACQUISITION - We have collected around 3000 images of leaf images of all the stage.

21. Cont... IMAGE PREPROCESSING RESIZING This method includes various techniques such as changing image size. The resize operation is performed it will convert the given image to dimension 640 X 640. Initial the pixel value of image is high after resizing the pixel value reduced by slight change in contrast of image. Original image Resized Image

22. CONT... GREYSCALE It is an image conversion technique in digital photography. It eliminates every form of colour information and only leaves different shades of gray. Original Image Greysacle Image

23. Cont... YOLOv8 Model Introducing Ultralytics YOLOv8, the latest version of the acclaimed real-time object detection, image segmentation and image classification model. YOLOv8 is built on cutting-edge advancements in deep learning and computer vision. Image classification models pretrained on the ImageNet dataset with an image resolution of 640. YOLOv8 automatically detects the objects and make bounding boxes around the object. After that it will send to CNN for training and testing the model.

24. Cont... CONVOLUTIONAL NEURAL NETWORK A convolutional neural network (CNN) is a type of deep neural network that is commonly used for image recognition, image classification, and other computer vision tasks it conatains 3 layers. a) Convolutional layer b) Pooling layer c) Fully connected layer

25. Cont... Convolutional layer The convolutional layer is the core building block of a CNN, and it is where the majority of computation occurs. It requires a few components, which are input data, a filter, and a feature map. Let’s assume that the input will be a color image, which is made up of a matrix of pixels in 3D. This means that the input will have three dimensions height, width, and depth which correspond to RGB in an image. We also have a feature detector, also known as a kernel or a filter, which will move across the receptive fields of the image, checking if the feature is present. This process is known as a convolution. The feature detector is a two-dimensional (2D) array of weights, which represents

26. Cont... Pooling layers Pooling layers, also known as down sampling, conducts dimensionality reduction, reducing the number of parameters in the input. Similar to the convolutional layer, the pooling operation sweeps a filter across the entire input, but the difference is that this filter does not have any weights. Instead, the kernel applies an aggregation function to the values within the receptive field, populating the output array.

27. Cont... Fully Connected Layers The name of the full-connected layer aptly describes itself. As mentioned earlier, the pixel values of the input image are not directly connected to the output layer in partially connected layers. However, in the fully-connected layer, each node in the output layer connects directly to a node in the previous layer. This layer performs the task of classification based on the features extracted through the previous layers and their different filters.

28. Cont... TRAINING AND TESTING The training data is the biggest (in -size) subset of the original dataset, which is used to train or fit the machine learning model. Firstly, the training data is fed to the ML algorithms, which lets them learn how to make predictions for the given task. Once we train the model with the training dataset, we can test the model with the test dataset

29. Cont... Training loss vs Testing loss on number of epoch for Bilva Patra leaf

30. Cont... Training loss vs Testing loss on number of epoch for other leaf

31. PERFORMANCE EVOLUTION Evaluate the trained model using the testing set. Measure performance metrics such as accuracy, precision, recall, and F1-score to assess the effectiveness of the model in leaf detection.Our developed model gives 92.37% accuracy.

32. FLOWCHART

33. USE CASE DIAGRAM

34. SEQUENCE DIAGRAM

35. Data Flow Diagram

36. Code Code for training the model

37. Cont... Code for testing the model

38. Cont... Code for testing the images given by user

39. Result Home Page for image selection

40. Cont... Shows selected image

41. Cont... Display result leaf is in initial stage

42. Cont... Display result leaf is in intermediate stage

43. Cont... Display result leaf is in final stage

44. Cont... Display leaf is bilva patra

45. Conclusion We have explored the process of vision-based identification of Bilva Patra leaf and their stages using Ultralytics YOLO deep learning model. We have covered the key steps involved in dataset preparation, model configuration, training, and testing. By following the provided code examples and guidelines, you can develop a robust system for identifying Bilva Patra leaves and categorizing them based on their stages. The Ultralytics YOLO framework provides a powerful and efficient solution for object detection tasks, including leaf identification.