Search Results

This paper describes the methods adopted to improve the image quality of the all-around view system. This system uses multiple vehicle-mounted cameras to capture images of the circumstances around the vehicle. The images undergo viewpoint transformation and are synthesized to create a single image seen from a virtual perspective overhead. The specific methods discussed here concern optimization of the design parameters for the camera orientation and a camera calibration method that does not involve physical movement of the cameras.

The aim of this study is to gain a better understanding of the mechanism of knocking with respect to flame propagation behavior based on 3D simulations conducted with the Universal Coherent Flamelet Model. Flame propagation behavior under the influence of in-cylinder flow was analyzed on the basis of the calculated results and experimental visualizations. Tumble and swirl flows were produced in the cylinder by inserting various baffle plates in the middle of the intake port. A comparison of the measured and calculated flame propagation behavior showed good agreement for various in-cylinder flow conditions. The results indicate that in-cylinder flow conditions vary the flame propagation shape from the initial combustion period and strongly influence the occurrence of knocking.

Forward Collision Alert (or Forward Collision Warning) systems provide alerts intended to assist drivers in avoiding or mitigating the harm caused by rear-end crashes. These systems currently use front-grille mounted, forward-looking radar devices as the primary sensor. In contrast, Lane Departure Warning (LDW) systems employ forward-looking cameras mounted behind the windshield to monitor lane markings ahead and warn drivers of unintended lane violations. The increasing imaging sensor resolution and processing capability of forward-looking cameras, as well recent important advances in machine vision algorithms, have pushed the state-of-the-art for camera-based features. Consequently, camera-based systems are emerging as a key crash avoidance system component in both a primary and supporting sensing role. There are currently no production vehicles with cameras used as the sole FCA sensing device.

In recent years, there has been a sustained effort by the automotive OEMs and suppliers to improve the vehicle driveline efficiency. This has been in response to customer demands for greater vehicle fuel economy and increasingly stringent government regulations. The automotive rear axle is one of the major sources of power loss in the driveline, and hence represents an area where power loss improvements can have a significant impact on overall vehicle fuel economy. Both the friction induced mechanical losses and the spin losses vary significantly with the operating temperature of the lubricant. Also, the preloads in the bearings can vary due to temperature fluctuations. The temperatures of the lubricant, the gear tooth contacting surfaces, and the bearing contact surfaces are critical to the overall axle performance in terms of power losses, fatigue life, and wear.

Liquid and vapor envelopes of sprays from a multi-hole fuel injector operating under closely-spaced double-injection conditions were investigated using a combination of high-speed schlieren and Mie scattering imaging. The effects of mass split ratio and dwell time between injections on liquid and vapor penetration have been investigated under engine-like pressures and temperatures. For the conditions evaluated, the results indicate that closely-spaced double-injection generally reduces liquid and vapor penetration.

Diesel engine exhaust aftertreatment components, especially the diesel particulate filter (DPF), are subject to various modes of degradation over their lifetimes. One particular adverse effect on the DPF is the significant rise in pressure drop due to the accumulation of engine lubricant-derived ash which coats the inlet channel walls effectively decreasing the permeability of the filter. The decreased permeability due to ash in the DPF can result in increased filter pressure drop and decreased fuel economy. A unique two-step approach, consisting of experimental measurements and direct numerical simulations using ultra-high resolution 3D imaging data, has been utilized in this study to better understand the effects of ash accumulation on engine aftertreatment component functionality.

The Engine-Out Particulate Matter (EOPM) was collected from a spark ignition engine operating in steady state using a heated quartz fiber filter. The samples were weighted to obtain an EOPMindex and were analyzed using Scanning Electron Microscopy. The EOP Mindex was not sensitive to the engine rpm and load. When the mixture is very rich (air equivalence ratio λ less than ∼ 0.7), the EOPM comprise mostly of soot particles from fuel combustion. In the lean to slightly rich region (0.8 < λ < 1.2), however, the EOPM are dominated by particles derived from the lubrication oil.

This paper describes an image processing method for recognizing white lines on the road ahead to accomplish automatic lane tracking. The behavior of detected points on the white lines can be regarded as multi-channel time-series data. The stochastic behavior of each point correlates with not only its own past behavior but also the behavior of other points around it. Behavior is identified by using a multi-variable auto-regressive (AR) model. This method can estimate the road configuration ahead logically even if the points cannot be detected clearly because a line is discontinuous or is hidden by a forward vehicle.

A method has been developed for improving the ease of viewing images on an in-vehicle display while a vehicle is moving and reducing carsickness. An attempt was made to mitigate carsickness by reducing sensory conflict by controlling the position of displayed images in synchronization with vehicle motions and passenger head motions produced by vehicle acceleration / deceleration forces. In the case of moving images, experimental results showed that, in addition to image position control, providing visual clues for distinguishing between the motions of the images themselves and the control motions can effectively reduce carsickness. The results further indicated that this method of controlling the position of displayed images is also effective in improving the ease of viewing images in a moving vehicle.

This paper presents a new method for calculating optical flow using data from a high frame-rate camera. We focused on a feature of image data captured with a high frame-rate camera in which objects do not move more than one pixel between successive frames. This approach eliminates repetitive processing for object identification among frames taken at different sampling times. High-speed processing hardware architecture was designed with sequential processing only, and the algorithm was implemented in a field programmable gate array. The resultant unit can calculate optical flow for a 640×120 pixel size image with a 480-Hz processing cycle and 0.5-μsec processing latency.

This paper describes a low-cost 48 × 48 element thermal imaging camera intended for use in measuring the temperature in a car interior for advanced air conditioning systems. The compact camera measures 46 × 46 × 60 mm. It operates under a program stored in the central processing unit and can measure the interior temperature distribution with an accuracy of ±0.7°C in range from 0 to 40°C. The camera includes a thermoelectric focal plane array (FPA) housed in a low-cost vacuum-sealed package. The FPA is fabricated with the conventional IC manufacturing process and micromachining technology. The chip is 6.5 × 6.5 mm in size and achieves high sensitivity of 4,300 V/W, which is higher than the performance reported for any other thermopile. This high performance has been achieved by optimizing the sensor's thermal isolation structure and a precisely patterned Au-black absorber that attains high infrared absorptivity of more than 90%.

The presence of liquid fuel inside the engine cylinder is believed to be a strong contributor to the high levels of hydrocarbon emissions from spark ignition (SI) engines during the warm-up period. Quantifying and determining the fate of the liquid fuel that enters the cylinder is the first step in understanding the process of emissions formation. This work uses planar laser induced fluorescence (PLIF) to visualize the liquid fuel present in the cylinder. The fluorescing compounds in indolene, and mixtures of iso-octane with dopants of different boiling points (acetone and 3-pentanone) were used to trace the behavior of different volatility components. Images were taken of three different planes through the engine intersecting the intake valve region. A closed valve fuel injection strategy was used, as this is the strategy most commonly used in practice. Background subtraction and masking were both performed to reduce the effect of any spurious fluorescence.

This paper describes the numerical and experimental approaches that were applied to study swirl injectors that are widely used in direct-injection gasoline engines. As the numerical approach, the fuel and air flow inside an injector was first analyzed by using a two-phase flow analysis method [VOF (Volume of Fluid) model]. A time-series analysis was made of the flow though the injector and also of the air cavity that forms at the nozzle and influences fuel atomization. The calculated results made clear the process from initial spray formation to liquid film formation. Spray droplet formation was then analyzed with the synthesized spheroid particle (SSP) method. As the experimental approach, in order to measure the cavity factor that represents the liquid film thickness, nozzle exit flow velocities were measured by particle image velocimetry (PIV).

This paper describes an image processing system for tracking a traffic lane by recognizing white lines on the road ahead. The system utilizes the features of the white lines and the Hough transformation to detect white line candidate points in images taken with a CCD camera. The parameters of the road configuration and vehicle attitude are estimated with an extended Kalman filter. This system has been applied to achieve a lane-keeping assistance system that provides steering control based on the host vehicle’s lateral position in its lane.

Computer simulation of extravehicular activity (EVA) is increasingly being used in planning and training for EVA. A space suit model is an important, but often overlooked, component of an EVA simulation. Because of the inherent difficulties in collecting angle and torque data for space suit joints in realistic conditions, little data exists on the torques that a space suit’s wearer must provide in order to move in the space suit. A joint angle and torque database was compiled on the Extravehicular Maneuvering Unit (EMU), with a novel measurement technique that used both human test subjects and an instrumented robot. Using data collected in the experiment, a hysteresis modeling technique was used to predict EMU joint torques from joint angular positions. The hysteresis model was then applied to EVA operations by mapping out the reach and work envelopes for the EMU.

The challenge of developing a robust, real-time driver gaze classification system is that it has to handle difficult edge cases that arise in real-world driving conditions: extreme lighting variations, eyeglass reflections, sunglasses and other occlusions. We propose a single-camera end-toend framework for classifying driver gaze into a discrete set of regions. This framework includes data collection, semi-automated annotation, offline classifier training, and an online real-time image processing pipeline that classifies the gaze region of the driver. We evaluate an implementation of each component on various subsets of a large onroad dataset. The key insight of our work is that robust driver gaze classification in real-world conditions is best approached by leveraging the power of supervised learning to generalize over the edge cases present in large annotated on-road datasets.

The engineering of electric propulsion systems requires time and cost efficient methodologies to determine system characteristics as well as potential component integration issues. A significant part of this analysis is the identification of the electromagnetic fields present in the propulsion system. Understanding of the electromagnetic fields during system operation is a significant design consideration due to the use of components that require large current(s) and high voltage(s) in the proximity of other control system items (such as sensors) that operate with low current(s) and voltage(s). Therefore, it is critical to quantify the electromagnetic fields produced by these components within the design and how they may interact with other system components. Often overlooked (and also extremely important) is an evaluation of how the overall system architecture can generate or react to electromagnetic fields (which may be a direct result of packaging approaches).

General Motors (GM) and the Virginia Tech Transportation Institute (VTTI) have partnered to conduct a series of studies characterizing the use and effectiveness of technologies designed to assist drivers while backing. A major emphasis of this research has been on Rear Vision Camera (RVC) systems that provide drivers with an enhanced view of the area behind the vehicle. RVC systems are intended to aid in positioning the vehicle when executing low-speed parking and backing-related tasks and are not necessarily well suited for detecting unexpected in-path obstacles (particularly if the RVC image is not coupled with object detection alerts issued to the driver).

In making driver workload assessments, it is important to evaluate the driver's level of brain activity because the operation of a motor vehicle presumably involves higher-order brain functions. Driving on narrow roads in particular probably imposes a load on the driver's brain functions because of the need to be cognizant of the tight space and to pay close attention to the surroundings. Test vehicles were fitted with a functional near-infrared spectroscopy (fNIRS) system for measuring bloodstream concentrations at 32 locations in the frontal lobe of the participating drivers in order to evaluate their levels of mental activity while driving on narrow roads. The results revealed significant increases in cerebral blood flow corresponding to the perceived workload. This suggests that increases in cerebral blood flow can be used as an effective index for estimating mental workloads.

High-speed (12 kHz) imaging of combustion luminosity (enhanced by using a sodium fuel additive) has been analyzed and compared to crank angle resolved heat release rates and mass fraction burn profiles in a spray-guided spark-ignited direct-injection (SG-SIDI) optical single-cylinder engine. The addition of a sodium-containing additive to gasoline greatly increases the combustion luminosity, which allows unintensified high-speed (12 kHz) imaging of early partially premixed flame kernel growth and overall flame propagation with excellent signal-to-noise ratio for hundreds of consecutive engine cycles. Ignition and early flame kernel growth are known to be key to understanding and eliminating poor burn cycles in SG-SIDI engines.