UNIT III OBJECT DETECTION USING MACHINE LEARNING
Object detection :
• Object Detection is a computer
vision technique to locate objects in
an image or in a video.
• Humans minds have been trained in
such a way that it can identify
various objects.
• The goal is to replicate this decision-
making intelligence using Machine
Learning and Deep Learning.
• Using a computer to mimic
this intelligence is the aim of
object detection.
UNIT III OBJECT DETECTION USING MACHINE LEARNING
Object detection:
• The coordinates of the objects in an image that an object detection system has
been trained to identify will be returned.
• Used in the image processing object detection like ADAS technology.(Advanced
Driver Assistance Systems (ADAS) are electronic technologies that help drivers
improve safety and comfort while driving)
• The object detection task combines subtasks of object localization and classification to
simultaneously estimate the location and type of object instances in one or more
images.
UNIT III OBJECT DETECTION USING MACHINE LEARNING
OBJECT DETECTION
ADVANTAGES OF OBJECT DETECTION
• Improved accuracy
• Automatic feature learning
• Analyse complex data
DISADVANTAGES OF OBJECT DETECTION
• Viewpoint variation
• Deformation
• Varying illumination conditions
OBJECT DETECTION MODELS
Two main approaches to creating object recognition
models:
• Creating and training an object detector
• Use a pre-trained object detector
Creating and training an object detector
• Have to build a network architecture that can learn the
characteristics of the items of interest in order to train a custom
object detector from scratch.
• To train the CNN, a sizable collection of labeled data must also be
assembled.
• CNN especially for the analysis and feature extraction from visual
data
• A personalized object detector can produce amazing outcomes.
• Configuring the layers and weights in the CNN manually takes a lot of
time and training data.
Use a pre-trained object detector
• Transfer learning is a technique that allows you to start
with a pre-trained network and then fine-tune it for
your application.
• It is used in many deep-learning object detection
procedures because the object detectors in this method
have previously been trained on dozens, if not
millions, of photos, it can yield speedier results.
USE CASES OF OBJECT DETECTION
1. Object Detection is the key intelligence behind autonomous
driving technology. It allows the users to detect the cars,
pedestrians, the background, motorbikes, and so on to
improve road safety.
2. We can detect objects in the hands of people, and the
solution can be used for security and monitoring purposes.
Surveillance systems can be made much more intelligent and
accurate. Crowd control systems can be made more
sophisticated, and the reaction time will be reduced.
USE CASES OF OBJECT DETECTION
1. Object Detection is also used for detecting objects in a shopping basket,
and it can be used by the retailers for the automated transactions. This
will speed up the overall process with less manual intervention.
2. Used in testing of mechanical systems and on manufacturing lines. We can
detect objects present on the products which might be contaminating
the product quality.
3. In the medical world, the identification of diseases by analyzing the
images of a body part will help in faster treatment of the diseases.
Object Detection methods: Machine Learning Methods
• 1. Image segmentation using simple attributes like shape, size, and
color of an object.
• 2. Aggregated channel feature (ACF), which is a variation of channel
features. ACF does not calculate the rectangular sums at various
locations or scales. Instead, it extracts features directly as pixel values.
• 3.Viola-Jones algorithm can be used for face detection. RANSAC
(random sample consensus), Haar feature–based cascade classifier,
SVM classification using HOG features.
Object Detection methods: Deep Learning Methods
1. R-CNN: Regions with CNN features
2. Fast R-CNN: A Fast Region–based Convolutional Neural Network.
3. Faster R-CNN.
4. Mask R-CNN: extends Faster R-CNN
5. YOLO: You Only Look Once architecture
6. SSD: Single Shot MultiBox Detector
OBJECT LOCALIZATION
• Object localization is a technique for determining the
location of specific objects in an image by
demarcating the object through a bounding box.
• Localization refers to finding the position of the
object in an image.
• In Image Localization, it means that the algorithm is
having a dual responsibility of classifying an image
as well as drawing a bounding box.
OBJECT LOCALIZATION & CLASSIFICATION
OBJECT CLASSIFICATION
• Image classification involves assigning a label or a class to an
entire image based on its content.
• The primary goal of image classification is to categorize an image
into one of several predefined classes or categories.
• This is achieved through the use of machine learning algorithms,
particularly convolutional neural networks (CNNs), which have
proven to be highly effective in extracting meaningful features from
images.
OBJECT CLASSIFICATION
• Image classification starts with preparing your data.
• In order to provide better image data for Computer Vision models to work with, this
process removes unwanted deformities and enhances some important parts of
the image.
• In essence, you’re preparing your data for processing by cleaning it for the AI
model.
• This phase is followed by object detection, where objects are localized, which
involves object segmentation and object position determination.
• Deep Learning algorithms then identify features and patterns in the image that can
be specific to a certain label. With the help of this dataset, the model gains
future accuracy improvements. In the final step, the ML algorithm divides
observed things into predefined classes using a suitable classification strategy. It
accomplishes this by contrasting desired patterns with picture patterns.
OBJECT CLASSIFICATION
OBJECT CLASSIFICATION
•Convolutional neural networks (CNN)- These deep learning
architectures are made especially for the analysis and feature
extraction from visual data. Convolutional, pooling and fully linked
layers are among the layers that make up a CNN.
•Transfer learning – This is an effective method that enhances image
classification performance by using pre-trained models. Transfer
learning makes training faster and more accurate by starting with a pre-
trained model and optimizing it on a particular dataset.
•Support vector machines (SVM) – Supervised learning models, or
SVMs, are excellent at classifying data into distinct groups. They are
appropriate for picture categorization because of their ability to manage
high-dimensional data.
