ABSTRACT
Armed banditry cases in Nigeria have risen drastically in recent years with an approximation of 200 cases in each month of 2020. Security forces in Nigeria have not completely repelled to the hilt these uprising crises due to a high number of cases, vast uninhabited areas across the country and low technological resources. This paper comprehensively discusses an intelligent camera for real time image recognition of armed bandits. Convolution Neural Network as a machine learning technique that models a system from a dataset has been adopted in accurate recognition of armed bandits. 5G networks that provide communications at fastest speed and low latency have been involved in integration of the intelligent cameras and signal transmission. The intelligent camera meshed with convolution neural network, speeded-up robust algorithms and 5G networks can localize, detect, and produce an alarm when guns are detected on images of a real-time video recording.
Author: Muhammad Alamin Ahmad
Key words: Armed bandits, Intelligent Camera, Machine Learning
Click Link Below For Full Paper, Questions and make your COMMENTS
INTELLIGENT CAMERA FOR REAL-TIME RECOGNITION
Can engineers resolve the armies banditry cases in nigeria
Thank you Aminu Ibrahim.
Well, Engineers have helped resolved myriads number of issues in developed countries. Great inventions using machine learning have proved to be exceptional in fighting modern problems facing not only Nigeria but the world at large.
Driving such inventions in Nigeria will make insecurity history.
Wow very interesting project Engr, Is this project operate with 5G network only and I need more explanation on how the camera can detect the banditry
Well you try, but is it possible for the camera to process person if he’s carrying small arm fire like pistols?
Thank you Nura Bashir Usman.
I really appreciate the question you asked.
The Models generated by machine learning depends on dataset. Any kinds of weapons of any shape (including piston) can be detected by the camera. A careful selection of data will make the system accurate and efficient.
You have done a Weldon job engineer. But how close will the camera will perform it’s desired function when the bandits are approaching ?
Very interesting idea and I hope this project may take our country to the next level, because our country need to be adopted to modern security system
My question here is on the detection techniques. only criminal guns it detect or all guns where captured including for the security personnel one
Thank you, Aliyu Bukar Usman.
Its indeed a challenge for the camera to distinguish police and bandits and the paper include that as a challenge.
But later advancements in techonology might tackle that problem.
This paper is undoubtedly a problem solver if really implemented, but the problem here is that will the system differentiate between a government personel weapon and a terrorist weapon when detected or will it send the same alarm? and also how reliable is this system if implemented?
In order for the system to distinguish police and bandits, a sophisticated machine algorithm which will be accurate in localization can be used to program the intelligent camera using data of police uniforms.
The intelligent camera will indeed be reliable without uncertainty but this depends on CNN and SURF algorithms discussed on the paper.
The system will undoubtedly respond when weapons are detected and in a matter of milliseconds, signals will be transmitted to a security base using 5G communications involved in the network.
This paper is outstanding, and it highlighted a big issue affecting Nigeria.
My question is how fast can the system send an alert when bandits are detected?
5G networks transmit signals at fastest speed, low latency and greater bandwidth. It helps in rapid machine to machine communication. If the camera detects weapons in few milli seconds all local authority would be alerted.
Good work Brother. This is the kind of work we want be seeing our young brilliant students doing. My question is how can you motivate our Government to implement this by installing CCTV cameras in our roads and public places so that crimes could be easily detected, traced and dealt with?
Motivating the government to finance and support such technologies requires much convincing strategy, the government should be notified that the solutions offered by such smart systems are efficient and accurate. This essay along with many other intellectual write up is a start to promoting awareness to battle insecurity in Nigeria.
Thank you.
So, do you mean that the camera can’t detect a suspect with out holding a gun? Thanks
Yes.
The system needs training data to model the network. The data can be of different variety and shapes (like guns, explosive, etc).
Thank u my brother ,,but my question is, can our country have the capability to install cctv cameras al over the nation. As we all knw that the arm bandits lives or camp them selves in villages or urban arears were no electricity and network services present. Can this be possible?
Nigeria is beyond capable in funding and supporting these technologies.
Considering the destructions insecurity have caused and the government expensive and archaic military mortar patrol in curbing the crises.
Thank you
That means you gonna need a wide area network topology for the cameras likewise a long-term power supply in order to keep them active since they are gonna be connected wirelessly to the network in order to have access to them. This is a very good and sensible idea.
Good job bro, I wish honourable Minister of communication see this, let our country moves forward, we are tired of all this bandits and bokoharam issues in the north, all u mentioned I’m convinced. No question.
Thank you brother.
Wao!!,very interesting ideal and a job well done to u my Eng.
My question is *like how much will the system cost*
The system will indeed cost in installment stage due to the requirements of infrastructure, power supply unit, 5G small base stations.
But operation cost and the reliability of the system would result be less costly considerably.
This is marvelous indeed. An effectual research coming at the right time, judging by the lingering cases we experience daily.
However, there is a huge challenge as regards the installation of these intelligent cameras where they may be needed more, without the attention of the bandits/criminals. Any practical solution in store?
It is indeed a challenge for the camera to be hidden but acknowledge the fact that the practical existing CCTV cameras had helped greatly in developed countries for example England the most sophisticated country with huge number of CCTV cameras where one can never hide from being seen.
The solutions that would be offered by intelligent camera for crime reduction will indeed enable Nigeria police to provide a forward step toward safety.
Very nice presentation young man, with doing this your effort has to be weight and credited
I really enjoy your ESSAY
Thank you Auwal for taking your time and reading this paper.
This is good idea engineer
But do you think insecurity in Nigeria will be overcome while 5G is not yet
5G is an emerging technology that promise a reliable future for intelligent technologies. If Nigeria would engage in 5G deployment more smart machines will leap forward and the solve the bugging problems affecting Nigeria
I actually admire this paper, the solution presented is surely unique and outstanding.
But I need more explanation on the camera and what differentiate from intelligent camera
Intelligent camera has the capability to capture and process images taken from a video recording. The camera is also capable of running specific algorithm on the image frames in order to produce desired signals.
Thank you.
A well written article, informative and a solution to the problems of insecurity
I believe if this work will be implemented the citizens of this country will no more words to utter than words of gratitude, because the paper highlited all the problems and presented a solution to them
Weldone Engineer.
More grease to your elbows.
Thank brother for taking your time and reading the paper. I believe if we can read and write about problems and solutions to issues facing Nigeria, we will surely extend the standard of our country to the highest possible acme of position.
I can’t help but notice that you mentioned something about video detecting guns on people, what if its not an actual gun don’t you think there will be too much false alarms.
If there is no gun the system will not produce any alert, and the accuracy of the system depends on number of dataset (100,000 or beyond data will be enough to model the network accurately).
Thank you.
QUESTIONS:
1. Will these cameras be installed in all primary and secondary schools all over the country?
2. How much will it take to make it to be enough for the 36states?
3. Is this camera only going to be installed in the northern parts of Nigeria only or all over?
4. Are there very qualified personnel’s that would be able to maintain and keep this equipment?
5. How would it functional?
6. Would this equipment create jobs for graduates all over the country?
1. Only if the school is insecure and susceptible to armed terrorist attacks.
2. It depends on the areas that needed protection.
3. Its to be installed across all vulnerable states in the country.
4. The camera is autonomous and it required no human intervention for detection.
5. It would function using Digital Image processing algorithm (CNN & SURF) for object detection in frames of video recordings. The algorithms will search for weapons on realtime human images. If weapons are detected the camera will send alert using rapid machine 2 machine communication (provided by 5G) to security personnel so that immediate response could be possible.
Interesting!
1.What are Different Types of Machine
Learning algorithms
2.How do we determine the machine learning algorithm used.
The types of learning algorithms depend on the data input. There is the supervised learning in which the data contains set of input and correct output. The other type of learning algorithm is the unsupervised learning in which the data consist of only raw data, in this case the algorithm looks for clusters in the data and tend to group them. However there is also another learning algorithm called transfer learning, this algorithm is a type of learning algorithm in which a pre-developed model (e.g, AlexNet, Resnet50 etc) are re-trained with the new data.
2. In my paper I proposed a CNN algorithm for training the camera on image detection of armed bandits. I did not however implement the project, this means only after implementation we can determined the algorithm. And still even after implementation to determine the algorithm we need to have a look at the source code written to build the model