Vehicle Safety Using Traffic Sign Detection and Object Detection
[1]Anisha Anchit, [2]Nancy Soni,[3] Nilesh Kumar,[4] Shradha Agarwal
Rajasthan Technical University, Poornima institute of Engineering & Technology
B.Tech Scholar, Deptt. of Information Technology
[1], [2],
[3],[4]
ABSTRACT
Due to carelessness of drivers while driving and violation of traffic rules, a large number of accidents occur today. According to latest road transport ministry report, a total of 4.97 lakhs road accidents were reported in 2011. To solve this problem, we propose to develop an automated system that would allow detecting traffic signs and object in front of the car along the way. To accomplish our aim, a traffic sign recognition program has been developed in MATLAB/Simulink environment and Sensors are used to detect the object.
Keywords: Automated System, Traffic Sign Detection, Object Detection,
Recognition, Matlab,Simulink Environment
1. Introduction
INTELLIGENT vehicles are becoming a part of our day to day life. Due to carelessness of drivers while driving and Violation of traffic rules, a large number of accidents occur today. Traffic signs carry essential information for drivers. They define prohibitions, give warnings, etc. Ignoring the presence of such signs can lead drivers into dangerous situations or even accidents. Automatic road sign detection and recognition systems can be used both to warn drivers in these situations and supply additional environmental information. In order to automate the process of traffic signs detection and recognition and alerting the driver about the traffic sign in traffic flow it`s necessary to solve a few problems
• analyse the systems existing at the moment
• justify the choice of the developed method and software
• build an info logical model;
• implement the selected option of the project
• make an experimental study
The main advantage of the automated system of traffic signs detection is a reduction of the risk imposed on the driver of the vehicle while driving.
Object detection adds to the development of automated system. Many accidents occur on the roads due to rash driving and ignoring objects in front. To detect the objects the system proposes to use sensors and alert the driver.
The paper is divided into the following sections: Section (1) contains the introduction, Section (2) contains related work, Section (3) contains Features of road sign, Section (4) contains Proposed scheme, Section(5) contains the methodology , Section (6) contains Experimental Results and Section (7) concludes the paper.
2. Related Work
A lot of research has been carried out for designing the automated systems with traffic sign recognition and object detection. Many authors use RGB colour space to identify traffic sign, studies the behaviour of RGB components of several road sign to sunset. Difference between any two components alone was considered for colour segmentation. The other colour spaces such as HSI, HSV, YIQ, YCbCr, CIExyz are available in literature.
Shape is another important characteristic of traffic sign that is used for segmentation. Varying techniques have been used for shape based segmentation such as canny edge detection, Hough transform, template matching, radial base symmetry and corner detection.
To recognize traffic signs, various methods for automatic traffic sign identification have been developed and show promising results.
The common approaches for these
1)neural network (NN)
2)nearest neighbour classification
3) support vector machine(SVM)
4) genetic algorithm (GA)
5) co-relation based pattern matching.
Different authors implement different techniques for object detection and many more methods have been evolved till date.
3. Features of road Sign
Traffic signs have been designed so that they are easily recognisable from natural and driving environment. The colour for traffic sign are chosen such that, it serves different purposes and is also distinguishable for the driver while driving. The signs are represented by fixed shapes like triangle; circle, octagon, and rectangle there are a number of traffic signs in India categorized as
(a)WARNING (40): A triangle with red coloured border and white background represents a warning sign. These signs alert the driver with hazard ahead
(b)COMPULSORY (27): Compulsory signs uses circle with red border and white background These signs restrict the action of drivers depending on the pictogram represented On the sign.
(c)REGULATORY (10): They are mandatory signs use to control the action of driver on the road. Blue circle with white border represents the regulatory sign.
(d)INFORMATORY (15): Important information like nearby hospitals, telephone booth, first aid, petrol pumps etc. come under this category. This information helps the driver in emergency in need. This sign are represented by White rectangle with thick blue border.
This makes a total of 92 traffic signs all together. These signs are mainly characterized by colour and shape. Figure 1 shows the different types of Indian traffic
Fig. 1 Different category of traffic signs
4. Proposed Scheme
This Project is expected to be implemented in vehicles as an integrated System. In our system, a camera would be mounted on the car and the input video signals would be sent to the matlab processing unit for detection and extraction of traffic sign. Sensors are used to detect the objects coming in front. The users of this system will be the driver who will have to get visual and audible warning when the traffic sign is detected or the objects are on the path of the vehicle.
5. Methodology
5.1 Traffic Sign Detection
Pattern recognition by video processing
Technique is taking video signals as an input and finding interested object. Then it has to be specified. Figure 2 shows our flow chart of traffic sign recognition system in Matlab/Simulink program.
Figure 2 Flow Chart of Program
5.1.1 Video Signal
Video signal are taken as input by a digital video camera mounted on the car, we have used image acquisition toolbox of Matlab/Simulink program to input video for processing.
5.1.2 RGB Colour Space Conversion
An RGB image involves three colours red, green and blue .The Intensity values of each pixels are in the range of 0-255. In our study, original video signals are converted to RGB colour space.
5.1.3 Pixel Labelling
A video signal involves several moving or
stationary objects. In general, traffic signs has fixed colours. In this study we primed to recognize the traffic signs in the video and this signs has specific colours. In a video frame, colour tone values of each pixel are determined by pixel labelling method. A matrix is established which is sum with video frame. After this part these two matrix is matched with desired values for each pixel. Following equations shows the algorithm of this process.
If (R>77) and (R-G>17) and (R-B>17) (1)
If (R>108) and (G>108) and (B>108) (2)
If (R<122) and (G<122) and (B<122) (3)
For this equations;
R: Represents the “red” in RGB color space.
G: Represents the “green” in RGB color space.
B: Represents the “blue” in RGB colour space.
Equation 1 shows the conditions for red pixel. If the conditions in the first equation are satisfied, this means the pixel colour is red. Equation 2 and 3 shows the condition of white and black respectively. In this part, pixel of video frame are labelled which are expressed with we defined as number. Also the other pixels in video frame are assigned fixed numerical values. In this way, the traffic sign is marked in video frame. All pixel values except red,
white and black are fixed to “0”.
5.1.4. Blob Analysis
Blob detection refers to visual modules that are aimed at detecting points and/or regions in the image that differ in properties like brightness or colour compared to the surrounding. In our system we are using the colour as a parameter to detect the required road sign.
Figure 3: Blob Detection
5.1.5 Comparison Objects
After making blob analysis, objects are
Detected by the blob is compared with template library of objects stored in the system. Dimension of the traffic sign is (90x90) pixel.
Figure 4: Stored templates
Figure 5: Detected road sign
Figure 4 shows the template database stored in the system ,when road sign comes in front of the car it is matched with the template database and shown on the window.
For comparing the road signs Template matching algorithm is used.
Template Matching
A template is a small image (sub-image)
The goal is to find occurrences of this template in a larger image That is, we want to find matches of this template in the image. For Example-
Figure 6: Template
Figure 7: Image
Here figure 6 shows the sub image to be compared and figure 7 shows the image in which the template is to be found.
5.1.6 Traffic Sign Recognition
If as a result of comparison matches with tolerance, object in video is labelled with name of traffic sign in template library.
5.2 Object detection
The basic concept of Ultrasonic sensor obstacle detection is to transmit Ultrasonic sensor the signal (radiation) in a direction anda signal is received at the Ultrasonic sensor receiver when the Ultrasonic sensor radiation bounces back from a surface of the object.
Figure 8: Basic of object detection
In our system we are using this basic concept for detection of object. As soon as the object is being detected a buzzer is sound to alert the driver.
Figure 9: Circuit Diagram
6. Experimental Results
6.1 Traffic sign detection
We use digital camera mounted on the car
ILLUMINATION TESTINGTC-No / Action To be Tested / Expected Result / Actual result
Detection Percentage / Recognition Percentage / Detection percentage / Recognition percentage
ILM_1 / In dark room / 100 / 100 / 0 / 0
ILM_2 / In low illumination / 100 / 100 / 50 / 10.5
ILM_3 / In medium illumination / 100 / 100 / 100 / 87.5
ILM_4 / In high illumination / 100 / 100 / 100 / 62.5
ILM_5 / In very high illumination / 100 / 100 / 0 / 10.5
in our experimental study to capture the video. Dimension of the video’s frame is 120x160 pixels. The video is recorded in AVI format because MATLAB/SIMULINK programme give best performance. The traffics sign are marked in video frames in a rectangular blob and are simultaneously matched with the templates saved in the template library. Template library consists of traffic sign in png format. In figure 9, a traffic sign which captured in video is compared with all the signs in the template library and the sign is displayed on the panel below.
Figure 9(a): Detection of road sign
.
Figure 9(b): Recognition and display of road sign
Test Results
a) Illumination testing
Figure 10(a): In low resolution
Figure 10(b): In Medium Resolution
DISTANCE TESTING IN MOVING CARTc_No / Distance Between Camera and Sign(in cm) / Distance of car when sign is detected(in cm) / Time taken(in sec) / Speed of car(in cm/s)
MC_1 / 70 / 50 / 10 / 5
MC_2 / 100 / 50 / 20 / 2.5
MC_3 / 130 / 50 / 20 / 2.5
MC_4 / 150 / 40 / 18 / 2.2
MC_4 / 170 / 35 / 18 / 1.94
Figure 10(c): In Very high resolution
b) Distance and speed testing in moving car
c) Distance and time testing in static car
DISTANCE TESTING IN STATIC CARTC_No / Distance Between Car and Sign / Detection / Recognition / Time Taken(in sec)
SC_1 / 0.05 / Yes / Yes / 2.97
SC_2 / 0.1 / Yes / Yes / 5.25
SC_3 / 0.15 / Yes / Yes / 5.29
SC_4 / 0.2 / Yes / Yes / 4.91
SC_5 / 0.25 / Yes / Yes / 6.49
SC_6 / 0.3 / Yes / Partially / 6.27
SC_7 / 0.35 / Yes / No / 15.95
SC_8 / 0.4 / Yes / No / 23.57
Figure 11(a): Distance is 30 cm
Figure 11(b): Distance is 40 cm
d) Miscellaneous testing
TEMPLATE SIZE TESTINGTC-No / Action to be tested / Expected Result / Actual Result
Detection Percentage / Recognition percentage / Detection Percentage / Recognition Percentage
TS_1 / With 2.5cmX2.5 cm / 100 / 100 / 100 / 62.5
TS_2 / With 7.5cmX7.5cm / 100 / 100 / 100 / 87.5
OTHER TESTING
TC_no / Action To Be Tested / Expected Result / Actual Result / Remark
OT_01 / When Object other than road sign is shown / No detection and recognition should be performed / No detection and recognition is performed / Pass
OT_02 / When Object with high intensity of red and black color is shown / No detection and recognition should be performed / Detection and recognition is performed / Fail
Figure 12: No detection other than road sign
6.2 Object Detection
Testing of SensorsParameters / IR / Ultrasonic Sensors
Range / 5 cm / 1m
Darkness / No Detection / Detects
In Fog / No Detection / Detects
7. Conclusion
In this Project, the related signs are accurately identified by a traffic signs
Determination program which is developed in Matlab/Simulink with a real time video taken by digital camera which includes variable traffic signs. The object is also detected by the sensors, and the alarm is generated. In the next phase the project is intended to determine more traffic sign in real time and real environment by increasing the size of the dataset. Also the objects can be categorised for detection.
References
[1] Musa AYDIN, Esra CALIK, Mustafa ISTANBULLU,”TRAFFIC SIGN RECOGNATION WITH VIDEO PROCESSING TECHNIQUE”, INTERNATIONAL JOURNAL OF ELECTRONICS; MECHANICAL and MECHATRONICS ENGINEERING Vol.2 Num.2 pp.(190-194)