Scribd
Upload
2 views

Leaf Detection System Machine Learning - Copy

The document outlines a project focused on developing a software application for plant leaf recognition using computer vision techniques. It describes the process of image acquisition, pre-processing, and classification using Convolutional Neural Networks (CNN) to identify plants based on leaf images. The system provides users with predicted results and relevant plant information through a graphical user interface.

Uploaded by

rincejohn80
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
2 views

Leaf Detection System Machine Learning - Copy

The document outlines a project focused on developing a software application for plant leaf recognition using computer vision techniques. It describes the process of image acquisition, pre-processing, and classification using Convolutional Neural Networks (CNN) to identify plants based on leaf images. The system provides users with predicted results and relevant plant information through a graphical user interface.

Uploaded by

rincejohn80
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 21

INTRODUCTION

 Detection and Recognition of various objects is one of the fascinating ability that human
possess.

 Human can identify the objects by extracting its various features like: shape, color, texture,
dimension etc.

 This way they can even recognize the object that possess slightly different properties than
they have ever seen.
 We have used the Computer Vision in the field of Plant Leaf Recognition.

 The purpose of this project is to develop a software that can identify the plant using
the captured image of its leaf.

 It can detect any plant leaf for which the system is trained. After detection and
recognition, it displays the various information of the plant that is helpful for the
users.
 Our system, Plant Leaf Recognizer undergoes series of steps to recognize the plant
accurately from the image of the leaf captured.

 First, the image is acquired, then it is pre-processed to remove noise and extract
features more clearly.

 At last we use Convolutional Neural Network ( CNN) for image classification and
the predicted result is displayed along with the details of the plant.
BLOCK DIAGRAM TRAINING
BLOCK DIAGRAM FOR TESTING
BASIC BLOCK DIAGRAM

Image Image Image Result


Acquisition Pre-processing Classification Prediction
IMAGE AQUISITION

 The first step of the system is image acquisition.

 The image of a leaf of the plant to be recognized is captured and fed to the system.

 Since our system is currently a desktop app, we have to locate the image via the user interface.

 The raw image is processed before it is sent to CNN.


IMAGE PRE-PROCESSING

 Pre-processing is used to get the outer shape of the image from the colored image of the leaf
and to remove any kind of external noises present in an image.

 The main idea of preprocessing is to enhance the image details so that features are clearly
found for further processing.

 Python Imaging Library (PIL), numpy and scikit-image was used to perform various image
preprocessing operations.
IMAGE PRE-PROCESSING
IMAGE CLASSIFICATION

 We used CNN for image classification.

 For the implementation of CNN, we have used keras library on top of tensorflow.

 CNN receives the image as a matrix of pixel values.

 A sequence of convolution, maxpooling and normalization is done in several


layers of CNN and is finally regularized.
IMAGE DATASETS

 Every image is a matrix of pixel values.

 The range of values that can be encoded in each pixel depends upon its bit size.

 Most commonly, we have 8 bit or 1 Byte-sized pixels.

 Thus the possible range of values a single pixel can represent is [0, 255].

 However, with coloured images, particularly RGB (Red, Green, Blue)-based images, the
presence of separate colour channels (3 in the case of RGB images) introduces an
additional ‘depth’ field to the data, making the input 3- dimensional.
IMAGE AS PIXEL MATRIX
RESULT PREDICTION

 This is the last stage of the system.

 At the end of the network, probability is calculated for each class and prediction is
made on the basis of probability.

 We have applied softmax function to convert the outputs to probability values for each
class.

 Then the class with the highest probability is chosen as the predicted output.
 The predicted result is then displayed in the Graphical User Interface (GUI) along
with the related details.

 The information regarding the plants are retrieved from the database.

 For the user interface, we have used a python library named Tkinter.
USECASE DIAGRAM
SCREENSHOTS

Home screen of the software


Dialog box after model is loaded
Dialog box after successful build of model
Window to locate training datasets
Window for selection of image to be recognized
Result display after successful prediction

You might also like