Search Results

Lane line detection is a very critical element for Advanced Driver Assistance Systems (ADAS). Although, there has been significant amount of research dedicated to the detection and localization of lane lines in the past decade, there is still a gap in the robustness of the implemented systems. A major challenge to the existing lane line detection algorithms stems from coping with bad weather conditions (e.g. rain, snow, fog, haze, etc.). Snow offers an especially challenging environment, where lane marks and road boundaries are completely covered by snow. In these scenarios, on-board sensors such as cameras, LiDAR, and radars are of very limited benefit. In this research, the focus is on solving the problem of improving robustness of lane line detection in adverse weather conditions, especially snow. A framework is proposed that relies on using Vehicle-to-Infrastructure (V2I) communication to access reference images stored in the cloud.

Bendability is a critical characteristic of sheet metal during the stamping process in automobile industry. Bending operation plays an important role in the panels forming of vehicles. In this study, the recently developed “Incremental Bending” method was utilized to evaluate the ambient bendability of 7xxx series avoiding bending crack. A 3D digital image correlation (DIC) measurement system is improved to capture the displacement and strain information on the stretched side of the sheet samples. The background, experimental method and data post-procedure are introduced in detail. After several sequential images acquisition and data processing, the major strain histories on the stretch zone of the samples are measured. With different bending process and parameters, the location of peak strain and the surface major strain distribution were evaluated as a function of R/T ratio (the inner radius over sheet thickness).

The assembling accuracy of two contactors during the relay switch production is an important factor affecting the quality of relay. An embedded machine vision quality Inspection system has been developed for electric relay production line inspection. The proposed system can provide online feedback on the quality of the relays by measuring the distance of the gap between the contacts of them. Two CMOS imaging sensors are operated for image acquisition and the parallel working mode is realized under dual-channel mode. A red light illumination system has been adopted to eliminate the imaging noise from the reflection of the surfaces of copper sheet. Before the test, the features areas in the image of same type relay is selected as template and saved in the computer. During the inspection procedure, a rotation invariance detection scheme based on circular projection matching algorithm has been used for fast recognizing and locating detected object with the help of these feature areas.

In this report, the DCX Stress Lab and the Tool Development & Test Support groups investigated automating a resonant bending crankshaft fatigue test. Fatigue testing, in general, is a laborious process since many samples are needed for analysis. This makes development cost and speed dependant on the component test efficiency. In the case of crankshaft resonant bending testing, both cost and speed are influenced by the manual feedback operation needed to run the current procedure. In order to increase the efficiency of this process, this project sought to automate the following tasks: maintaining the load on the part, reacting to resonance changes in the part, mapping resonance changes, logging the number of cycles, and discerning resonance frequency shift failure modes objectively.

Current engine development processes typically involve extensive steady-state and simple transient testing in order to characterize the engine's fuel consumption, emissions, and performance based on several controllable inputs such as throttle, spark advance, and EGR. Steady-state and simple transient testing using idealistic load conditions alone, however, is no longer sufficient to meet powertrain development schedule requirements. Mapping and calibration of an engine under transient operation has become critically important. And, independent engine development utilizing accelerated techniques is becoming more attractive. In order to thoroughly calibrate new engines in accelerated fashion and under realistic transient conditions, more advanced testing is necessary.

Fatigue testing is a specialized form of mechanical testing that is performed by applying cyclic loading to a coupon or structure. Two common forms of fatigue testing are load controlled high cycle and strain controlled low cycle fatigue. Some strain measurement device, such as extensometers, strain gage, that are often used as a feedback sensor on strain controlled fatigue test. However, in applications where strain controlled fatigue testing could face some extreme conditions as well as high temperature and unusual sizing which requires the strain measurement to be nondestructive and full field. While digital image correlation (DIC), an advanced optical measurement technique, has a decent solution on challenges of fatigue testing measurement. The problem is how to turn DIC from a measurement system to a feedback controller unit. Due to the developments in camera and computation techniques, the sequential process can now be performed as a parallel process.

A multi-sensor Digital Image Correlation (DIC) system is employed to measure the deformation of metal specimens during tensile tests. The multi-sensor DIC system is capable of providing high quality contour and deformation data of a 3D object. Methodology and advantages of the multi-sensor DIC system is introduced. Tests have been done on steel and aluminum specimens to prove the performance of the system. With the help of the multi-sensor DIC system, we proposed our approaches to determine the forming limit based on shape change around the necking area instead of calculate the FLD based on the in-plane strains. With the employed system, all measurements are done post-deformation, no testing controlling mechanism, such as load force control or touching control, is required. The extracted data is analyzed and the result shows a possibility that we may be able to improve current technique for Forming Limit Diagram (FLD) measurement.

This work presents a method to estimate the human body orientation using 2D images from a person view; the challenge comes from the variety of human body poses and appearances. The method utilizes OpenPose neural network as a human pose detector module and depth sensing module. The modules work together to extract the body orientation from 2D stereo images. OpenPose is proven to be efficient in detecting human body joints, defined by COCO dataset, OpenPose can detect the visible body joints without being affected by backgrounds or other challenging factors. Adding the depth data for each point can produce rich information to the process of 3D construction for the detected humans. This 3D point’s setup can tell more about the body orientation and walking direction for example. The depth module used in this work is the ZED camera stereo system which uses CUDA for high performance depth computation.

In this research, ball-on-plate reciprocating sliding wear tests were utilized on austempered and quench-tempered gray cast iron samples with and without shot-peening treatment. The wear volume loss of the gray cast iron samples with different heat treatment designs was compared under equivalent hardness. The phase transformation in the matrix was studied using metallurgical evaluation and hardness measurement. It was found that thin needle-like ferrite became coarse gradually with increasing austempering temperature and was converted into feather-like shape when using the austempering temperatures of 399°C (750°F). The residual stress on the surface and sub-surface before and after shot-peening treatment was analyzed using x-ray diffraction. Compressive residual stress was produced after shot-peening treatment and showed an increasing trend with austempering temperature.

In this work, a multi-camera Digital Image Correlation (DIC) system is applied to measure the material properties of aluminum (5754) specimens. Such tests are usually done using 2D (one-camera) or 3D (two-camera) DIC systems. A multi-camera DIC system includes three or more cameras and inherits all the advantages of a conventional 3D DIC system (with two cameras) such as, full-field measurement, high accuracy and high speed. In addition, this system further improves the measured results by including redundant data. In this work, we will show the potential of this system to measure a variety of material properties at one time.

In 1998, Rouhana et al. described development of a new device, called the IR-TRACC (InfraRed - Telescoping Rod for Assessment of Chest Compression). In its original concept, the IR-TRACC uses two infrared LEDs inside of a telescoping rod to measure deflection. One LED serves as a light transmitter and the other as a light receiver. The output from the receiver LED is converted to a linear function of chest compression using an analog circuit. Tests have been performed with IR-TRACC units at various labs around the world since 1998. A first-generation IR-TRACC system was retrofit into a Q3 dummy by TNO. Similarly, a mid sized male Hybrid III dummy thorax and a small female Hybrid III dummy thorax have been designed by First Technology Safety Systems (FTSS) such that each contains 4 second-generation IR-TRACC units. The second-generation IR-TRACC is the result of continued development by FTSS, especially in the areas of the analysis circuit, manufacturing and calibration methods.

In the development of several outside mirror designs for vehicles, a high frequency noise (whistling) phenomenon was experienced. First impression was that this might be due to another source on the vehicle (such as water management channels) or a cavity noise; however, upon further investigation the source was found to be the mirror housing. This “laminar whistle” is related to the separation of a laminar boundary layer near the trailing edges of the mirror housing. When there is a free stream impingement on the mirror housing, the boundary layer starts out as laminar, but as the boundary layer travels from the impingement point, distance, speed, and roughness combine to trigger the transition turbulent. However, when the transition is not complete, pressure fluctuations can cause rapidly changing flow patterns that sound like a whistle to the observer. Because the laminar boundary layer has very little energy, it does not allow the flow to stay attached on curved surfaces.

Digital Image Correlation (DIC) technology is a powerful tool in the field of experimental mechanics to obtain the full-field deformation/strain information of an object. It has been rapidly applied in industry in recent years. However, for the large-angle full-field strain measurement of small-sized cylindrical objects, it’s still a challenge to the DIC accurate measurement due to its small size and curved surface. In this paper, a measurement method based on the multi-camera DIC system is proposed to study the compressive performance of small-sized cylindrical materials. Three cameras form two stereo DIC measurement systems (1 and 2 cameras, and 2 and 3 cameras), each of which measures a part of the object. By calibrating three cameras at the same time, two stereos DIC coordinate systems can be unified to one coordinate system. Then match the two sets of DIC measurement data together to achieve large-angle measurement of the cylindrical surface.

As vacuum suction cups are widely used in stamping plants, it becomes urgent and important to understand their performance and failure mode. Vacuum suction cups are employed to lift, move, and place sheet metal instead of human hands. Occasionally the vacuum cups would fail and drop parts, even it would cause expensive delays in the production line. In this research, several types of vacuum cups have been studies and compared experimentally. A new tensile device and test method was developed to measure the pulling force and deformation of vacuum cups. The digital image correlation technique has been adopted to capture and analyze the contour, deformation and strain of the cups under different working conditions. The experimental results revealed that the relevant influential parameters include cup type, pulling force angles, vacuum levels, sheet metal curvatures, etc.

This paper introduces an industrial application of digital image correlation technique on the measurement of aluminum edge stretching limit. In this study, notch-shape aluminum coupons with three different pre-strain conditions are tested. The edge stretching is proceeded by standard MTS machine. A dual-camera 3D Digital Image Correlation (DIC) system is used for the full field measurement of strain distribution in the thickness direction. Selected air brush is utilized to form a random distributed speckle pattern on the edge of sheet metal. A pair of special optical lens systems are used to observe the small measurement edge area. From the test results, it demonstrate that refer to the notched coupon thickness, pre-tension does not affect the fracture limit; refer to the virgin sheet thickness, the average edge stretch thinning limits show a consistent increasing trend as the pre-stretch strain increased.

It is commonly known that in an automotive manufacturing assembly line several workers perform either a common task or a number of different tasks simultaneously, and there is a need to represent such a multi-worker operation realistically in a digital environment. In the past years, most digital human modeling applications were limited only in a single worker case. This paper presents how to simulate multi-worker operations in a digital workcell. To establish an effective communication and interaction between the mannequins, some existing commercial software package has provided a digital input/output mechanism. The motion for each mannequin is often programmed independently, but can be interrupted anytime by the other digital human models or devices via a communication channel.

This paper proposes a direct nanometer measurement method using the technique of high magnification diffraction imaging. The principles of employing both gratings and speckles are described, in which the illumination distances that generate high magnified and focused images are discussed along with demonstration test results. In addition, the potential extension to sub-nanometer displacement measurement is discussed.

This work presents a method for estimating human body orientation using a combination of convolutional neural network (CNN) and stereo camera in real time. The approach uses the CNN model to predict certain human body keypoints then transforms these points into a 3D space using the stereo vision system to estimate the body orientations. The CNN module is trained to estimate the shoulders, the neck and the nose positions, detecting of three points is required to confirm human detection and provided enough data to translate the points into 3D space.

In this paper, Digital Image Correlation Method (DICM) is employed to measure the shear mechanical property of the new style automotive structural adhesive specimens and traditional spot welded specimens under quasi static uniaxial shear tensile test. This experiment adopts a non-contact measuring method to measure the strain of specimens. A CCD and a computer image processing system are used to capture and record the real-time surface images of the specimens before and after deformation. Digital correlation software is used to process the imagines before and after deformation to obtain the specimen's strain of the moment. And then both the force-displacement curve and the stress-strain curve during the tensile process could be obtained. The test and analysis results show that the new style structural adhesive specimens have a great advantage with the spot welded specimens. It provides experimental evidence for further improvement of this structural adhesive.

The edge fracture occurs more frequently during the forming procedure by using the material with higher strength. To avoid the edge fracture that happens during the manufacturing, the edge cracking limit at different pre-strain level needs to be determined. The edge of the part under forming is conventionally manufactured by mechanical cutting, and the edge cracking limit under this circumstance is already heavily studied. In recent years, laser cutting is more applied in the automotive industry to cutting the edge due to the following advantages over mechanical cutting: easier work holding, higher precision, no wearing, smaller heat-affected zone, etc. The change cutting method could lead to a different behavior to the edge cracking limit at different pre-strain level. In this paper, the edge cracking limits of sets of pre-strained coupons with different pre-strain levels are tested. Half of them is cut by the conventional punch method, and the other half uses laser cutting.