Object Detection Using Deep Learning

Batch Number: A12

by,
Under the guidance of,
1. 3VC17CS013: B.GANESH
Prof. R P RAJESHWARI,
2. 3VC17CS021:CHETANKUMAR.U
Assistant Professor,
3. 3VC17CS015:B.LINGRAJ
Department of CSE,
4. 3VC17CS028: GURURAJ.D
RYMEC
 Contents
• Introduction
• Area and Problem statement
• Objective
• Benefits to the society/Organization ..etc
• Methodology
• System Architecture
• Partial Demonstration
• Expected Results
• References

Department .of Computer Science, RYMEC, Ballari 2

 1. Introduction
•Object detection, one of the most fundamental and
challenging problems in computer vision, seeks to
locate object instances from a large number of
predefined categories in natural images.
•Deep learning techniques have emerged as a
powerful strategy for learning feature representations
directly from data and have led to remarkable
breakthroughs in the field of generic object
detection.
• As a longstanding, fundamental and challenging
problem in computer vision, object detection has
been an active area of research for several decades

Department .of Computer Science, RYMEC, Ballari 3

 2. Area and Problem statement
Area
• Deep Learning
Problem Statement
Object detection is customarily considered to be much harder
than image classification, particularly because of these five
challenges: dual priorities, speed, multiple scales, limited data,
and class imbalance.
Sloution:
The field of computer vision has experienced substantial
progress recently, owing largely to advances in deep learning,
specifically convolutional neural networks (CNNs). Image
classification, where a computer classifies or assigns labels to an
image based on its content, can often see great results simply by
leveraging pre-trained neural nets and fine-tuning the last few
throughput layers.
Department .of Computer Science, RYMEC, Ballari 4
 3. Objective

•The goal of object detection is to detect all instances

of objects from a known class, such as people, cars or
faces in an image.

•The main objective of the Object detection is it

allows us to identify and locate objects in an image or
video by using is a computer vision technique.

• With this kind of identification and localization,

object detection can be used to count objects in a scene
and determine and track their precise locations, all
while accurately labeling them
Department .of Computer Science, RYMEC, Ballari 5
4. Benefits to the society/Organization ..etc
•Object detection can be used in various industries and
there is basically no restriction to the way you use it – it
all depends on the possible value the technology can
bring to the customers
•E-commerce - Image search, Product recommendation.
•Security - We can use the object recognition
technology to identify any suspicious activity within the
area of your company and get an immediate
notification about it.
•Healthcare - One of the most important areas for using
object recognition in medicine is scanning of MRIs and
CT scans with the aim to identify a disease or its early
symptoms
Department .of Computer Science, RYMEC, Ballari 6
•Self-driving cars - Self-driving cars are nothing new but
there are still many issues related to this trend, with the
primary one being safety. Autonomous cars need to react
fast enough to avoid any dangerous situations and that’s
where image recognition steps in and helps resolve the
issue.

•Summing up - Object detection among the technologies

that shape-shift modern business development and set its
direction.
Dept.of CSE, RYMEC 7
5. Methodology

Department .of Computer Science, RYMEC, Ballari 8

•Download and explore the dataset-
• Importing (or installing) Keras, Tensorflow , Open CV and other
packages on system
• Load dataset from disk
• Create training and testing splits
• Create model
• Compile and save the model
• Train the model on training data (split)
• Evaluate the model on test data (split)
• Prediction on new data using trained model

9
6. System Architecture

Department .of Computer Science, RYMEC, Ballari 10

 In the proposed procedure, CNNs are applied to detect and classify lung cancer
CT scans of the patients collected from hospitals. Convolution Neural Networks is a
sort of deep learning paradigm applied for processing data which has a grid pattern
like images; it is all about using Deep Learning with Computer Vision. A good way to
gain foreknowledge about this technique is to imagine a Neural Network Architecture
also how it is practiced to visual tasks i.e. Video and Images. Furthermore, the
Convolution Neural Networks is an important technique used for Object Recognition;
create Facial Recognition, Self-Driving Cars. A Convolution Neural Network is a
Deep Learning algorithm that can take in image as input, with assigning importance
learnable weights and biases to various objects inside this image and be capable of
differentiating one from the other. In addition, the preprocessing required for this
technique is much lower if comparing with other classification algorithms. The role of
the CNN is for reducing the images without losing features that are important for
getting a good prediction. A typical CNN consists of three types of operation layers:
the convolution layer (CONV), the pooling layer (POOL), and finally the classifier
layer (FC), as exemplified in the figure. 11
7. Partial Demonstration

Department .of Computer Science, RYMEC, Ballari 12

 8. Expected Results

Fig 1: Potted Plant Fig 2: CellPhone

Dept.of CSE, RYMEC 13
Fig 3: Vase Fig 4 : Bird 14
 9. References
[1] D. Erhan, C. Szegedy, A. Toshev, et al, “Scalable object detection
using deep neural networks,” 2014 IEEE Conference on Computer
Vision and Pattern Recognition, 2014, pp. 2155-2162.
[2] A. Borji, M. M. Cheng, H. Jiang, et al, “Salient object detection: A
benchmark,” IEEE Transactions on Image Processing, vol. 24, Dec
2015, pp. 5706-5722.
[3] Ce Li and Yachao and V Zhang and Object detectors in serving
robots. IEEE International Conference on Systems – 2018
[4] Malay Shah and Rupal Kapdi , “Object Detection Using Deep
Neural Networks” International Conference on Intelligent Computing
and Control Systems ICICCS 2017.

Department .of Computer Science, RYMEC, Ballari 15

THANK YOU

Dept.of CSE, RYMEC 16