Object Classification System in Low Visibility Driving Conditions
In low visibility conditions (such as in night), the likelihood of a vehicle crash increases. In this situation, one can?t determine what is in the front of the vehicle and if it is approaching to or moving far away and if collision can happen. In the 2nd case, if there is
2025-06-28 16:28:41 - Adil Khan
Object Classification System in Low Visibility Driving Conditions
Project Area of Specialization Electrical/Electronic EngineeringProject SummaryIn low visibility conditions (such as in night), the likelihood of a vehicle crash increases. In this situation, one can’t determine what is in the front of the vehicle and if it is approaching to or moving far away and if collision can happen.
In the 2nd case, if there is a collision and the passengers injured, they may suffer till helped by someone outside the vehicle, until they don’t reach to the right nearest hospital.
- Thus, we directed to this issue by developing a system presented as 'Object Classification System' to aid driver in low illuminations by making the most relative image of outside object in front of driver upon LCD in order to prevent vehicle collision.
- Moreover, a system will be designed for detecting vehicle accident, detecting the crash type and giving post-accident measures immediately.
Our projects' main focus comprises of the following areas:-
1-To detect the object
2-To classify the object
3-Make decisions on the basis of object classification 4-To detect accident
5-Make decisions on the basis of vehicle clash
- For object classification system, LiDAR is interfaced with raspberry pi.
- For providing post-accident measures, MEMS MPU6050 sensor and GPS/GSM module is interfaced with PIC microcontroller.The reason for using MCU to interface sensor for the system providing post-accident measures is to make a separate modern PCB for this system.
We are determined to make the overall system efficient and cost-effective.
Project ObjectivesThe main objectives of our project are:
Primary Scope
- To develop an algorithm to detect a vehicle and distinguish it with other objects.
- To develop an algorithm to classify the vehicle
- To implement the proposed algorithm
Secondary Scope
- To detect accident and crash type (collision and rollover)
- To provide post-accident measures
Since the project is Object Classification System in Low Visibility Conditions so the central goal is as follows:
- First of all, Object presence in a particular vicinity is detected by LIDAR(Light Detection and Ranging), read the data values by Raspberry Pi, calculating its distance, comparing with the pre-set values and if meeting the specified criteria utilizing the mapped data (taken by Lidar) and Image Processing to display the most related image corresponding to the object on LCD.
Secondary goal of system initiates after the vehicle collision, whose key purpose is to detect accident and provide medical aid as soon as vehicle crashes.
- To accomplish this, PIC microcontroller is interfaced with MEMS sensor specifically accelerometer and gyroscope (integrated on a chip known as MPU6050) along with GSM/GPS module SIM808.
- PIC Microcontroller is interfaced with MPU6050 which process accelerometer data to measure changes in acceleration forces along the 2-axis(x and y) for vehicle’s collision detection and gyroscope data to measure changes in rotation angles (roll and pitch) for detecting vehicle rollover. As soon as the crash detects, the processing involved in PIC MCU for the purpose of crash detection by comparing the data values continuously obtained from sensors with the preset threshold values, sim808 sensor comes into action, in which GPS and GSM both are integrated, GPS calculates the vehicle location and finds the nearest hospitals while GSM reports to that opted hospitals and responds back to victim with a request approval.
Since many of the vehicle collisions’ causes corresponds to low visibility conditions, which can be avoided via 'Object Classification System' (Project's Primary Scope)
- The precious lives and the valuables can be made safe from accidents in low light conditions by developing the closest image of outside nearby vehicles on LCD corresponding to the vehicle’s lane.
Secondly, if accident happens and if it is severe then it takes an indefinite amount of time to take appropriate measures and to provide medical assistance to the injured passengers.
- So comes our 'Secondary Project Goal' which is to detect accident and providing post-accident measures depending upon the accident severity.
The main role of our project is to prevent deaths and injuries by reducing the number of car accidents and the serious impact of those that cannot be avoided. Accident prevention and detection is the project keystone.
Technical Details of Final Deliverable- Our final deliverable will detect a vehicle coming in near proximity and classify it by mapping the object and computing its dimensions.
Object recognition using a LIDAR sensor attached to the raspberry pi module, using AI/ machine learning domain and implementing convolutional neural network (CNN) break down the images into 2D frames and train the algorithm to identify various objects in the surrounding.
- Post-accident measures will also be provided, if the vehicle collides or gets rolled over[Detected by MPU 6050 and crash type identified by pic microcontroller], the location coordinates[using GPS] will be sent[using GSM] to the nearby hospitals if no response is observed by anyone in the vehicle. All these features will be designed using modern PCB techniques designing along with efficient coding.Thus, the post-accident measure system is designed on the Multilayered PCB.
Providing post-accident measures in case of accident and giving the real-time location of the sufferer to nearby hospitals using a GPS module that is programmed to communicate with the satellite.
Our end product will be a small sized kit integrated with the above features mentioned, compact design, multi-functional (Object Classification / Post Accident measures), cost-effective and is eco-friendly.
Final Deliverable of the Project HW/SW integrated systemCore Industry TransportationOther Industries Others , Security Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Good Health and Well-Being for People, Affordable and Clean Energy, Decent Work and Economic Growth, Life on LandRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 45370 | |||
| Lidar | Equipment | 1 | 12000 | 12000 |
| Raspberry Pi 3 | Equipment | 1 | 13000 | 13000 |
| Micro SD Card | Equipment | 1 | 1000 | 1000 |
| c-type Charger | Equipment | 1 | 700 | 700 |
| Ethernet Cable | Equipment | 1 | 150 | 150 |
| LCD Screen | Equipment | 1 | 3000 | 3000 |
| HDMI Cable | Equipment | 1 | 350 | 350 |
| Raspberry Pi Case | Miscellaneous | 1 | 400 | 400 |
| PIC microcontroller | Equipment | 1 | 2000 | 2000 |
| MPU 6050 sensor | Equipment | 1 | 350 | 350 |
| Crystal Oscillator | Equipment | 1 | 20 | 20 |
| programmer Pickit | Equipment | 1 | 2400 | 2400 |
| PCB (fabrication) | Equipment | 1 | 4000 | 4000 |
| resistors | Equipment | 20 | 5 | 100 |
| capacitors | Equipment | 20 | 10 | 200 |
| Soldering Iron,wire,paste,stand | Miscellaneous | 1 | 700 | 700 |
| GSM/GPS module | Equipment | 1 | 2000 | 2000 |
| Project Banner | Miscellaneous | 1 | 1000 | 1000 |
| Case/Kit | Miscellaneous | 1 | 2000 | 2000 |