Scribd
Upload
17 views10 pages

Cat Dog Classification Project

This project implements a binary classification system to distinguish between cats and dogs using the ResNet50 architecture, involving image preprocessing, segmentation, feature extraction, and machine learning classification. The dataset consists of 1,000 images from Kaggle's Dogs vs. Cats dataset, processed through various techniques to enhance quality and isolate animals. The model's performance is evaluated using metrics such as accuracy, precision, recall, and F1-score, demonstrating the effectiveness of the approach.

Uploaded by

Anwar Chouchane
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
17 views10 pages

Cat Dog Classification Project

This project implements a binary classification system to distinguish between cats and dogs using the ResNet50 architecture, involving image preprocessing, segmentation, feature extraction, and machine learning classification. The dataset consists of 1,000 images from Kaggle's Dogs vs. Cats dataset, processed through various techniques to enhance quality and isolate animals. The model's performance is evaluated using metrics such as accuracy, precision, recall, and F1-score, demonstrating the effectiveness of the approach.

Uploaded by

Anwar Chouchane
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/ 10

Cat/Dog Image Classification

Using ResNet50
A Binary Classification Project with
Image Preprocessing and Machine
Learning
Project Overview
• This project aims to implement a complete
image processing pipeline for binary
classification (cat vs. dog) using ResNet50
architecture and various techniques such as
image preprocessing, segmentation, feature
extraction, and machine learning
classification.
Dataset
• The dataset used is Kaggle's Dogs vs. Cats
dataset, containing 1,000 images (500 cats and
500 dogs) with a standardized resolution of
512x512 pixels. The dataset is split into three
subsets: training, validation, and testing.
Image Preprocessing
• In this stage, image quality is enhanced by
performing the following tasks:
• 1. Resizing images to 128x128 pixels.
• 2. Normalizing pixel values (scaling between 0
and 1).
• 3. Reducing noise using filters.
• 4. Optionally converting to grayscale.
• 5. Contrast enhancement using histogram
equalization.
Image Segmentation
• Segmentation isolates regions of interest
(animals). Techniques include:
• 1. Classical Methods: Thresholding, Edge
Detection (Canny, Sobel), Color Segmentation.
• 2. Advanced Methods: DeepLabV3+, U-Net,
Mask R-CNN.
• Segmentation creates binary masks for cats
and dogs, helping isolate animals from the
background.
Feature Extraction
• In feature extraction, we calculate geometric,
textural, and structural features:
• 1. Geometric: Area, perimeter, aspect ratio,
circularity.
• 2. Texture: Gray-Level Co-occurrence Matrix
(GLCM), Local Binary Patterns (LBP).
• 3. Structural: Hu moments, color histograms.
• These features help describe the images for
classification.
Machine Learning Classification
• The images are classified using machine
learning algorithms:
• 1. Support Vector Machine (SVM)
• 2. k-Nearest Neighbors (k-NN)
• 3. Random Forest
• 4. Logistic Regression
• The model is evaluated using a confusion
matrix and accuracy metrics.
Code Walkthrough
• The code includes the following key steps:
• 1. **Dataset Loading & Preprocessing:**
Using `cv2` and `Pandas` to load and clean the
dataset.
• 2. **Data Augmentation & Generators:**
Creating `ImageDataGenerator` for training,
validation, and test sets.
• 3. **Model Creation:** Using ResNet50 with
customizable frozen layers.
• 4. **Model Training:** Applying early
Model Evaluation & Results
• After training, the model's performance is
evaluated on the test set.
• We use metrics like accuracy, precision, recall,
and F1-score to compare the models.
• Results are displayed using confusion matrices
and performance curves.
Conclusion
• This project demonstrates the use of image
preprocessing, segmentation, feature
extraction, and machine learning classification
for the task of distinguishing between cats and
dogs.
• ResNet50, a powerful pre-trained model, was
fine-tuned for the task, providing high
performance.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy