This project is to find out the dynamic threshold value for the traffic lane extraction
At the early stage, we found that the threshold value of the image was the main item we had to solve. This is because the threshold value of the image can provide us the better image on the traffic lane extraction.
Dynamic threshold value determination is to produce a continuous threshold value to the traffic lane extraction.
By using the LabVIEW Program, we can easily extract the light intensity and HSI data of the image and get the threshold value by comparing the data with the threshold value database.
Table of Contents
No. Description Page
A. Introduction…………………………………………………………. 3
B. Literature Search………………………………………………………………….. 3
C. Design Concept……………………………………………………………. 4
D. Block Diagram………………………………………………………. 4
E. Experimental Record ………………………………………………………………… 5
F. Project Timeline…………………………………………………………………… 6
G. Result and Conclusion……………………………………………………………
H. References……………………………………………………………………. 7
Since the 21st Century, Autonomous vehicles are the hit topics in the world which is very important on the Intelligent Transport System. The basic provision of the intelligent vehicles should provide the information of route direction, sense objects or pedestrians; prevent impending collisions and lane departure warning to the driver.
The benefits of autonomous vehicles are:
a. Reduce traffic collisions – autonomous vehicles increase reliability and quick respond time compare with human.
b. Fewer traffic congestion and enlarge roadway capacity – the safety gaps between vehicles area are reduced and better traffic flow management.
c. Higher speed limit.
... middle of paper ...
...ngbill, David Lieb and Sebastian Thrun, “Optical Flow Approaches for Self-supervised Learning in Autonomous Mobile Robot Navigation”, Autonomous Navigation in Dynamic Environments, 2007
 Shengyan Zhou, Yanhua Jiang, “ A Novel Lane Detection based on Geometrical Model and Gabor Filter”,2010 IEEE Intelligent Vehicles Symposium, pp.59-64, Jun, 2010
 J.M. A’ lvarez and A. Lo’pez, “Novel Index for Objective Evaluation of Road Detection Algorithms”, In Proc. of the IEEE Intelligent Transportation System, Oct. 2008.
 Yue Wang, Dinggang Shen and Eam Khwang Teoh, “Lane Detection Using Catmull-Rom Spline”, In Proc. of the IEEE IntelligentVehicles, 1998
 Tsung-Ying Sun, Member, IEEE, Shang-Jeng Tsai and Vincent Chan, “HSI Color Model Based Lane-Marking Detection”, In Proc. of the IEEE Intelligent Transportation System Conference, pp. 1168-1172, September, 2006
In this essay, the author
Explains that this project is to find out the dynamic threshold value for the traffic lane extraction.
Explains that autonomous vehicles are the hit topics in the 21st century, which is very important on the intelligent transport system. the basic provision of intelligent vehicles should provide the information of route direction, sense objects or pedestrians, and prevent impending collisions.
Explains that autonomous vehicles reduce traffic collisions and increase reliability and quick response time.
Explains that less traffic congestion and enlarge roadway capacity reduce safety gaps between vehicles area and better traffic flow management.
Explains that the vehicles can park far away after drop off the passenger and return back as request by passenger.
Explains that lane and road, elimination of moving car by comparing the shape and size of the object, and.
Opines that lane-mark edges are used to detect a car's lanes, but this method relies on clear marks.
Explains that the detection images are transform to the lane models and compute the lanes. the complexity of computations for the transformation of models lowers the real time capability.
Explains that moving vehicles will eliminate by comparing the size and shape of the object.
Explains that lane coherence verification is the final step of line extraction, whereby the previously detected boundary width and orientation are stored.
Defines the region of interest (roi) in order to eliminate the un-used object.
Cites hsu-yung cheng, bor-shenn, pei-ting tseng, and kuo-chin fan in ieee transactions on the intelligent transportation system.
Cites claudio rosito jung and christian roberto keller in image and vision computing 23: 1192-1202, 2005.
Presents wei liu, hongliang zhang, bobo duan, huai yuan and hong zhao's "vision-based real-time lane marking detection and tracking", in ieee intelligent transportation system.
Presents andrew lookingbill, david lieb, and sebastian thrun's "optical flow approaches for self-supervised learning in autonomous mobile robot navigation".
Presents shengyan zhou and yanhua jiang's paper, "a novel lane detection based on geometrical model and gabor filter", ieee intelligent vehicles symposium.
Presents j.m. a’ lvarez and a. lo’pez's novel index for objective evaluation of road detecting algorithms.
Presents yue wang, dinggang shen, and eam khwang teoh, "lane detection using catmull-rom spline", in ieee intelligentvehicles, 1998.
Explains that tsung-ying sun, ieee member, shang-jeng tai, and vincent chan, "hsi color model based lane-marking detection", ieee intelligent transportation system conference.
Explains that at the early stage, the threshold value of the image was the main item they had to solve, because it can provide us the better image on the traffic lane extraction.
Explains the mechanism of lane departure warning system, which is designed to prevent the cost of accidents by driver error, distraction and drowsiness.
Explains the advantages of his color model-based lane-marking detection.
Explains that the early stage of the project is to develop the traffic lane detection and reconstruction system and provide the necessary information to the driver.
Further research can be done on increasing the reliability of the autonomous car system, like safety enhancements, what must the car do when there is a roadblock – (implement a alternative routing algorithm) or how to handle lane changes or encounter a diversion on the road, these can be numerous other issues. These are all hot research areas of now and many ideas are coming up and implemented. But as per this topic is considered a solution MUST be found to all these problems.
In this essay, the author
Explains that the paper discusses various issues to be covered while implementing a gps based autonomous car.
Explains that modern cell-based gps systems use network provided gps coordinates to track the car on the globe.
Explains that the arduino microcontroller is good but not suitable to handle real world problems on its own. an android device is backed by qualcomm snapdragontm s4 pro.
Explains that the earth's radius is not a perfect sphere, but decreases as we move towards the magnetic poles. the solution is to integrate the above listed problems in the math calculations.
Opines that applying autonomous vehicle technology in real world requires gps signalling and sensor technology, which brings more flexibility to meet real traffic situations.
Researchers from Intelligent Vehicle Initiative, a partnershipd between the U.S. Department of Transportation and automobile industries are making advances on systems which, if successful, can tell you if you're about to run off the road or hit another car. The system is a pretty complicated construction of lasers which are set all over the vehicle. Since there are always blindspots, the idea is that the lasers compare the path of your car to the paths and speeds of the vehicles around you. If something is your car's path, collision radar systems can warn you and calculate the distance you have to stop. A camera is even programmed to look for painted lines on the road to determind if you are off center or drifting into another lane. Some advanced systems will even be able to detect lines when snow is covering the ground.
In this essay, the author
Opines that the increasing role of technology in the development of cars shows its potential with these in-the-works safety features.
Explains that night vision is a system that uses infrared cameras to help drivers see much farther distances at night while driving.
Explains that collision radar is a complicated construction of lasers which compare the path of your car to the paths and speeds of the vehicles around you.
Explains that driver's attention is always the most crucial safety measure, but the more they know about what is going on around them, the better they'll be able to drive. collision radar and night vision are both examples of how technology is being used to increase safety in cars.
Autonomous Cars or the Self Driving Cars are capable to drive themselves without any human intervention. The Autonomous cars are designed such that they can drive themselves by sensing their environment and navigate without human input. Autonomous cars sense their surroundings using technologies like RADAR, LIDAR, GPS and Coy computer vision. Some autonomous vehicles update their maps based on sensory input, allowing the vehicles to keep track of their position even when conditions change or when they enter uncharted environments. According to the National Highway Traffic Safety Administration (NHTSA), United States has classified these self driving cars in some levels. For example, Level 0- The driver completely controls the vehicle at all times. Level 1-Automated braking and electronic stability control of the vehicle are automated. Level2- Two controls can be automated in combination like cruise control along with lane keeping. Similarly the other levels have advanced controls, like car senses conditions when it requires the driver to take control, in another level the car performs all the safety critical functions for the entire trip.
In this essay, the author
Explains that autonomous cars are designed to drive themselves without human intervention. they sense their surroundings using technologies like radar, lidar and gps, and update their maps based on sensory input.
Explains why self-autonomous cars are needed, and how they can be used to get around los angeles.
Explains that millions of people have died or became disabled in traffic accidents in the last ten to 15 years. most of the accidents are caused by human error and according to statistics, the number of lives lost each year will likely double.
Explains the purpose of the project, which is to create an autonomous car which can drive itself automatically. it can reduce distance between cars, reduce traffic jams, avoid human error, and allow disabled people to drive.
Describes the technologies that will make self-autonomous cars successful. traffic lanes recognition system provides assistance in locating the positions of the lane boundaries on the road.
Explains that lidar is becoming more and more popular as a guidance system for autonomous vehicles. the 32 lasers in the lidar system on ford's fusion scan the road and its surroundings 20,000 times per second.
Explains that the current autonomous car problem is caused by using a very expensive 3d lidar ($ 75,000), with incredibly high resolution.
Explains that the solution is a minimal 3d lidar with 1 m pixel/s that would only cost $4000 and 3 special cameras mounted to recognize from images the marker lines, borders and real time position of the car.
Explains the 6 electronic parts of the 3d lidar: base microcontroller, driving stepper – lidar, interface for rs232, fast convertor analogical digital spi, wireless transmitter.
Describes the lane detection system, which recognizes traffic lanes with three different web cams placed on the left and right side of the car and one in front at the driver’s seat.
Explains that one of the most promising features in autonomous cars is fcw (forward collision warning). it uses cameras to calculate the distance between your and other car.
Describes a blind spot detecting system that detects the area behind your car and the car next to.
Explains lane departure warning (ldw) technology, which uses camera and radar to detect lane markings, alerts the driver if the car drifts away from its lanes.
Explains that traffic sign recognition (tsr) is used to regulate traffic signs, warn a driver, and command or prohibit certain actions.
Explains that traffic signs from the common database have a property called feedback.
Explains that traffic signs detection is based on color filtration, cascaded multi-layer perception neural networks, and a common traffic sign database using gps coordinates and google maps.
Explains how google's software scans and reads traffic signals and sends that information to the car. they explain how a mounted camera would be the easiest and most logical solution.
Explains how the software works by first receiving images taken by the camera, filtering images, classifying images by searching for red, yellow, and green blobs and labeling them accordingly.
Concludes that the material below can be used for conclusion or further research. parallel parking is a challenge for many drivers.
Explains that bosch is working on a smart-phone app that will allow the driver to get out of the car, which will park itself, although he’ll need to hold down the button on his cell phone while monitoring the parking.
Explains that nhtsa is moving toward mandating that vehicles come equipped with wireless systems termed "v2v" that would allow cars to "talk" to each other, exchanging speed and location data so that they can avoid accidents.
Explains that the study discussed traffic sign recognition (tsr) using neural networks technique. the first strategy is to reduce the number of mlp inputs by pre-processing the traffic sign image.
Explains the network architecture which reduced complexity by selecting a suitable error criterion for training. the cross-validation technique was implemented with training data, validating data set, and test set.
Imagine that your car could detect hazardous roads before you could sense it? A feature that provides this type of protective driving exists within European cars. This valuable safety feature that is offered in European cars i...
In this essay, the author
Explains that technology and the automotive industry go hand in hand with constant improvement to components of cars. american and european car companies provide their buyers with innovative variety and revolutionary luxuries.
Explains the importance of safety within cars is a growing desire among all car buyers, which is why night vision capability is an important feature to european cars.
Explains that the active braking system is a valuable safety feature that adds safety to european cars such as bmw, mercedes-benz, and audi.
Explains that blind spot detection is an important feature which provides guaranteed safety to drivers all over the world.
Explains that european entertainment features are more advanced and sophisticated than american car companies. audi has integrated ipod connectivity and voice commend.
Explains that bmw's idrive is the easiest and most intuitive way to control entertainment, information, communication, and navigation functions.
Compares the mercedes-benz s550's selectable cabin fragrance with the american car industry.
Explains that european car companies produce and offer vehicles which are far superior to its american competitors.
Explains that bmw efficientdynamics and jeff daniels look under the skin of today's cars.
Analyzes why american cars are considered so bad when compared to european ones.
Further development of simulation programs have enhanced the traffic engineering field to provide in depth analysis of current, real world transportation networks. These simulation guides analysis into better interpretation of these networks in order to make informed decisions towards design additions and improvements for a given network. Specifically, microscopic simulation provides even greater detail for analyzing various traffic scenarios and implementing ITS strategies to improve the performance of a network. This paper summarizes different research studies conducted towards effective implementation of ITS strategies into real world scenarios through microscopic simulation modeling. More specifically, freeway control ITS strategies will be covered in detail.
In this essay, the author
Summarizes research studies conducted towards effective implementation of its strategies into real world scenarios through microscopic simulation modeling.
Explains that microscopic simulation provides further access into its procedures and studies since the majority of applications are technologically and iteratively focused.
Explains how henry liu's research case study analyzed a congested freeway network in southern california. its strategies were implemented individually and in different combinations to observe the effect on the network.
Explains how its strategies centered on highway control benefit transportation networks with extensive highway usage. moshe ben-akiva's study looked into managing traffic congestion using highway access control, route control and lane control with mitsim.
Cites liu, henry, and ben-akiva, moshe, for their case study on its applications.
The behavioral models of the traffic as a system has been developed and studied for several decades and the core behavioral models, Gipps’ car-following/acceleration model and lane-changing model are more or less enough to predict behavioral outcomes of vehicles in homogeneous traffic conditions with lane discipline. In mixed traffic conditions with weak/no lane discipline however, the behavior of a vehicle cannot be defined separately using the car-following and lane-changing models. The behavior of a vehicle under such conditions is very dynamic – depending on the behavior of other vehicles in its vicinity. Also, there exists a continuous interaction among the vehicles and therefore, every vehicle is influenced by its surrounding vehicles.
In this essay, the author
Explains that gipps' car-following/acceleration model and lane-changing model predict behavioral outcomes of vehicles in homogeneous traffic conditions, but the behavior of a vehicle under such conditions is dynamic.
In conclusion, with the careful implementation of these cameras coupled with engineering measures, injuries and fatalities can be reduced at signalized intersections. Technology is expanding at a rapid pace and some of it will benefit the traffic signal industry. New inventions are already being talked about in the media. Soon we may see the human factor eliminated as autonomous cars take the roads in pursuit of an accident free world.
In this essay, the author
Analyzes how a chicago tribune article reported that 574 motorists have been ticketed for running red lights since 2007 in the chicagoland area.
Explains that they suffered shoulder injuries that took two years to recover with surgery to their shoulder.
Concludes that with the careful implementation of these cameras coupled with engineering measures, injuries and fatalities can be reduced at signalized intersections. technology is expanding at a rapid pace and will benefit the traffic signal industry.
Explains that red-light-running camera enforcement is a relatively new technology, which was implemented in the 1990s.
Explains how the first application of traffic control was a police officer standing at the intersection to prevent collisions by directing traffic from different directions. the manual of uniform traffic control devices (mutcd) came to life in 1935.
States that the institute of transportation engineers (ite) has developed the formula for intersection yellow/amber clearance timings. the fhwa claimed the use of cameras was a factor in reducing the number of red light runners.
Explains that the fhwa describes two different categories of drivers that run red lights: intentional and unintentional.
Explains that the florida department of highway safety and motor vehicles reported that red-light cameras reduced crashes for the second year in a row.
Opines that red-light-runners put everyone's life in danger without an officer present at an intersection.
...ngo from an innovative engineering idea, to a fairly effective method of controlling traffic flow. As of today, there are currently 600 intersections that are in need of a useful method to improve conditions regarding traffic congestion, a plan to help prevent accidents thus reducing fatalities, and relieve drivers of unscrupulous police officers. For that reason, two twin robots are being tested now in two busy intersections in Kinshasa to analyze if in fact the number of collisions, fatal accidents, and the overall traffic flow is reduced. If the road networks in Kinshasa can achieve a level of service D, C (target), or even a B, as stated in the North American highway standards from AASHTO and Highway Capacity Manual then the roads will increase traffic volume, capacity, and flow. Nevertheless, the African countries have other transportation and traffic standards
In this essay, the author
Explains the importance of traffic control measures in the united states. kinshasa, the capital of the democratic republic of congo, in africa, uses two robots to ease traffic flow.
Explains that therese ir isay kirongozi and her team of engineers were interested in a project that would help traffic because kinshasa has disordered driving and traffic congestions.
Explains that the traffic robots are 8 feet tall and have a solar panel to provide energy, which is advantageous for districts that have minimal electrical power.
Explains that before the two robots were put at two different locations each, the poor city police officers earned minimal salaries and for that reason are often seeking compensation from commuters.
Explains that traffic collisions have diminished at the two intersections monitored 24 hours a day, 7 days of the week by the robots.
Explains that engineers theresa and her team created robots because this method can prosper in other countries that would like to invest in traffic robotics.
Explains that traffic robots have been great news in the republic of congo from an innovative engineering idea, to a fairly effective method of controlling traffic flow.
This article written by Chris De Santis explains the three major traffic problems that are impossible to solve because of our human driving differences, but points out that the self-driving cars of the future have the clear solutions. The 1st problem is the ripple effect. When one car reacts to something on the highway and slows down, this starts a ripple effect that will cause a car to stop eventually down the line. The effect will keep going, keeping the traffic slow. The solution is self-driving cars that are programmed to stay equidistance in the middle between 2 cars and accelerate simultaneously. The 2nd problem are the slow reaction times of drivers that cause grid lock at the next
In this essay, the author
Explains the three major traffic problems that are impossible to solve because of our human driving differences, but points out that the self-driving cars of the future have the clear solutions.
J. N.Kapur, P. K Sahoo and A. K. C Wong, “A New Method for Picture Thresholding Using the Entropy of the Histogram, Computer Vision, Graphics and Image Processing,” Vol. 29, No. 3, pp. 273-285, 2007.
In this essay, the author
Introduces a ghost free high dynamic range imaging algorithm for obtaining ghost-free high dynamic range (hdr) images. multiple image fusion based hdr methods work only on condition that there is no camera and object movement.
Explains the importance of high spatial resolution, true color reproduction, and high dynamic range (hdr) for real world scenes acquisition.
Explains that multiple image-based hdr algorithms require high computational complexity and long processing time, which makes consumer applications impossible.
Explains the proposed hdr algorithm that consists of histogram separation method, stretching, edge-preserving denoising, and ldr image fusion.
Explains the proposed algorithm uses the concept of genetic algorithm and entropy maximization technique as a histogram separation method to specify stretching region that also reduce the loss of information in the final hdr image.
Proposes a histogram separation method that uses the concept of entropy based histogramme separation and genetic algorithm to specify stretching regions.
Explains that in kapur image histogram separation algorithm, we consider gray level entropy of two subsets.
Explains that h0 calculates the entropy of the first subset that has 0 to gray levels, and h1 calculate the gray level that maximizes equation (3).
Explains that they employ a genetic algorithm for finding the gray level point that maximizes the objective function.
Explains that genetic algorithm determines the gray level point that maximizes equation (3), which is the fitness function for genetic algorithm.
Explains that si(x, y) is the pixel intensity in subset image i. the stretched histogram is then transformed back to the rgb color space.
Explains that the proposed method employs an edge-preserving denoising algorithm that takes detailed high-frequency regions from a noisy ldr image with an appropriate amount of weighting between them.
Explains that g is the output of averaging filter, and is weighting factor.
Explains that mobile image devices have limited available memory and computational power, which makes multi-frame hdr imaging difficult. ldr image fusion is needed to combine the three images that are generated by the proposed algorithm.
Explains that hc (i, j) represents the hdr image, and i_ck (k u, n, o), represent contrast stretched ldr images simulating under-, normally- and over-exposed images. the weighting factor is computed using the gaussian-shaped function.
Proposes a single image-based ghost-free hdr imaging algorithm using histogram separation method and edge preserving denoising technique.
Describes g. ward, e. reinhard, s. pattanaik, and p. debevec, high dynamic range imaging, morgan kaufmann publisher, 2005.
Cites y. bandoh, g. qiu, m. okuda, s. daly, t. aach, and o. au. recent advances in high dynamic range imaging technology.
Explains p. debevec and j. malik, “recovering high dynamic range radiance maps from photographs,” proc.
Analyzes e. land's "retinex theory of color vision," scientific american, vol. 237, pp.
Explains how image fusion and motion stabilization can be used to capture still images in high dynamic range scenes.
Explains that motion-blur free camera system splitting exposure time, ieee trans. consumer electronics, vol. 54, no. 3, august 2008.
Explains wu, xie, rahardja, and li's robust and fast anti-ghosting algorithm for high dynamic range imaging.
Cites j. im, s. lee, and paik, “improved elastic registration for ghost artifact free high dynamic range imaging,” ieee trans. consumer electronics, vol.
Presents bilcu, burian, knuutila, and vehvilainen's paper, "high dynamic rangeimaging on mobile devices," ieee conf.
Describes j. n.kapur, p. k sahoo, and a. c. wong's methods for picture thresholding using the entropy of the histogram, computer vision, graphics and image processing.
Explains r. a. hummel, "image enhancement by histogram transformation, computer graphics and image processing," vol.6, no.2, pp.184-195, 1977.
Explains that real-time image restoration for digital multifocusing in a multiple color-filter aperture camera, optical engineering, vol. 49, no. 4, april 2010.
Explains how single image-based ghost-free high dynamic range imaging uses local histogram stretching and spatial-adaptive denoising.
Explains kapur's algorithm for image histogram separation into two subsets based on the entropy of the gray level.
Explains how the genetic algorithm generates an initial population of size 64 by randomly selecting the gray level point of an image. the fitness function f() = h_0+h_1 is evaluated.
Explains how they split image histogram dataset d into two subsets d0 and d1.
Explains that the proposed algorithm transforms the rgb color space to the hue, saturation, and value (hsv), and stretches the histogram of the v channel in the following manner.
Illustrates how the histogram stretched image with amplified noise and the color-mapped edge image are processed by the denoising algorithm given in equation (6).
Explains the proposed algorithm that generates ghost free hdr image using the single input image. three sets of differently exposed ldr images acquired by different exposure values (evs) in the dynamic environment produced ghost.
Explains how the proposed algorithm generates hdr images using three ldr images that are under-, normal- and over-exposed, and compares with the existing algorithm.
Compares the mean squared error (mse), peak signal to noise ratio (psnr), normalized cross correlation (ncc) and normalized absolute error (nae) values of the proposed algorithm with existing methods.