DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

R18 B.TECH(C.S.E) IV YEAR I SEMESTER - B SECTION

ADMISSION BATCH INDUSTRIAL ORIENTED MINI PROJECT ACADEMIC YEAR

Batch Number : 04

Title of the Project : SONAR ROCK Vs MINE PREDICTION

Project Work doing at : Industry/Pantech E-Learning/Conscience/Own

Team members Details(First Person is Project Leader):

01. 20E31A0546 S PUSHPANJALI [email protected] 8919140234

Details of Project Guide in the college:

01. CH.NARSI REDDY Assistant Professor 7729901871

Details of Project Supervisor at Industry:

Signature of the Signature of the Signature of the

Underwater Mine usage by the Naval Defense System provides great Security but also possesses a threat
to the marine life and submarine vessels as the mines can be easily mistaken for rocks. So, we need a much
more accurate system to predict the object as it is very dangerous if a mistake is made. To have a great
accuracy we need more accurate data to generate accurate results. Our idea presents a method for prediction
of underwater mines and rocks using Sonar Signals. Sonar Signals are used to record the various
frequencies of underwater objects at 60 different angles. We constructed three binary classifier models
according to their accuracy. Then, prediction models are used to predict the mine and rock categories.
Python and Supervised Machine Learning Classification algorithms are used construct these
prediction models.

II. OBJECTIVE:

Underwater mines are a strategic military tool to protect any country‘s naval borders. These mines consist
of an explosive material, sensing device and a fuse mechanism that triggers an explosion when ships or
other submarines come in contact with it. This was the scenario during previous generations, but the newly
developed mines, on the contrary, are equipped with advanced technology, such that their sophisticated
sensors even detect some combinations of acoustic and magnetic signals. There are some other mines that
are equipped with artificial intelligence to detect any false signals that trigger the explosion. Several mines
have been implanted in the oceans during World War II and even today , so that countries protect their
naval borders from enemies. They pose an extreme threat to ships and submarine sand thus must be detected
and eliminated.

SONAR technique is commonly used to detect these mines. SONAR, which stands for sound navigation
and ranging, is a sound propagation technology used primarily for underwater communication, navigation,
and/or object identification. The SONAR returns are then used to identify the objects the waves were
reflected from. But, it is hard to correctly determine if the target identified is a mine or a rock. The dataset
used contains SONAR signals reflected off of a metal cylinder and a cylindrically shaped rock positioned
on a sandy ocean floor. The impinging pulse was a wideband linear FM chirp (ka = 55.6), and both targets
were around 5 ft (1.52 m) long. Returns were gathered over a distance of 10 meters from the cylinder at
90° aspect angles and from the rock at 180° aspect angles

III. SCOPE:

The objective of this study is to develop a highly accurate prediction system for distinguishing between
underwater mines and rocks using Sonar signals. While underwater mines serve as a crucial component of
naval defense systems, their presence poses a significant threat to marine life and submarine vessels,
especially due to the potential misidentification of mines as rocks. Therefore, there is a pressing need to
develop an advanced system capable of accurately classifying underwater objects to ensure the safety and
security of marine environments. To achieve this objective, the study utilizes a dataset provided by
Gorman, R.P., and Sejnowski, T.J. (1988) that is specifically designed for training machine learning
models. The data set consists of Sonar signals capturing the frequencies of underwater objects from60
different angles. By leveraging this data, the study aims to construct three binary classifier models, each
tailored to maximize accuracy in differentiating between mines and rocks. Python and supervised machine
learning classification algorithms are employed to build these prediction models. The study explores
various classification algorithms, selecting the most suitable ones that yield the highest accuracy in
prediction. By harnessing the power of supervised machine learning techniques, the objective is to create
robust and reliable models capable of accurately classifying Sonar signals and providing real-time
predictions on the nature of the underwater objects encountered. By achieving a high level of accuracy in
predicting underwater mines versus rocks, the proposed system aims to significantly enhance the safety
and efficiency of marine operations.

IV. REVIEW OF LITERATURE:

Elakkiya, Jincys, Rejinan, and Tamilmalar in Classification of SONAR targets using advanced neural
networks discuss classification of SONAR targets into rocks and mines using Meta-Cognitive Neural
Network (MCNN) and Extreme Learning Machine (ELM) classifiers to achieve an acceptable efficiency.
It further elaborates on the strategies given by MCNN like sample delete strategy, neuron growth strategy
etc and advantages of ELM due to no tuning required for the hidden layer. After analyzing the performance
of MCNN and ELM, the paper concludes that the training efficiency of MCNN is 81.7% and testing
efficiency is 87.5% while that of ELM is 88% and 84% respectively. Abdul-Qader in Techniques for
classification: rocks vs mines . It is concerned with a procedure for separating metal cylinders (mines) from
objects having cylindrical shapes (rocks) utilizing SONAR signals and the three methods listed below: NN,
ANFIS and kNN. Use of sequential forward selection method makes three techniques give a better
accuracy. KNN gives a better accuracy of 93.27%. ANFIS gives the best accuracy of 100%. NN gives a
better accuracy of 88.42%.Jongwon Seok in Active Sonar Target Classification Using Multi-Aspect
Sensing and Deep Belief Networks. For active SONAR target classification to enhance classification
performance, a multi-aspect-based sensing system is provided. In order to create 3-dimensional highlight
models, the active SONAR returns from targets are synthesized using the ray tracing algorithm. To extract
the features, FrFT (Fractional Fourier transform) is applied to SONAR returns. With the FrFT-based
features, four different targets are classified using deep belief networks. The proposed feature extraction
method showed better performance than the conventional neural network classifier. Shaik Firasat Ali and
Abdul Rasool MD in Sonar data classification using multilayer perceptron), brief about classification of
SONAR data with the help of Deep learning using a multilayer perceptron (and the perceptron is being
used as a linear classifier). The model generated here runs for 1000 epochs, where in each epoch the whole
input data is used to train the model. Each epoch's cost, mean square error, and training accuracy are
calculated and shown. With an epoch count of 999, the model can accurately predict whether the object
might be a rock or a mine with an accuracy of roughly 85%.

HARDWARE REQUIREMENTS:

1. OS-Windows 11(64 bit)

V. SOFTWARE REQUIREMENTS:

1.PYTHON
 2.MACHINE LEARNING

VI. REFERENCES

[1] “Analysis of hidden units in Layered Network Trained to Classify Sonar Targets” in Neural
Networks, Vol.1,pp.75-89(1988).

[2] “Connectionist Bench (Sonar, Mines vs Rocks).” Connectionist Bench (Sonar, Mines vs
Rocks)|Kaggle,

[3] Shantanu, et al. “Underwater Mines Detection Using Neural Network”. Underwater Mines
Detection Using Neural Networks-

[4] “Underwater Mine Detection Using Symbolic Pattern analysis of Sidescan Sonar Images.”
Underwater Mine Detection Using Symbolic Pattern Analysis of Sidescan Sonar Images,
ieeexplore.ieee.org,https://ieeexplore.iee.org/document/5160102.

[5] “A Review of Underwater Mine Detection and classification in Sonar Imagery” Stanislaw
Hozyn ORCiD-0000-0003-1422-0330.

[6] “A Study on Detection and Classification of Underwater Mines using Neural Networks”

N.Geethalakshmi, P.Subashini, S.Ramya ISSN:2231-2307.

[7] “Underwater Mine Detection using Image Processing” Abhishek, Arjun, Bharathesh, Kavitha
Prof. Manonmani Dr.Shanta Rangaswamy e-2395-00