A Network That Balances Accuracy and Efficiency for Lane Detection

In the automatic lane-keeping system (ALKS), the vehicle must stably and accurately detect the boundary of its current lane for precise positioning. At present, the detection accuracy of the lane algorithm based on deep learning has a greater leap than that of the traditional algorithm, and it can achieve better recognition results for corners and occlusion situations. However, mainstream algorithms are difficult to balance between accuracy and efficiency. In response to this situation, we propose a single-step method that directly outputs lane shape model parameters. This method uses MobileNet v2 and spatial CNN (SCNN) to construct a network to quickly extract lane features and learn global context information. Then, through depth polynomial regression, a polynomial representing each lane mark in the image is output. Finally, the proposed method was verified in the TuSimple dataset. Compared with existing algorithms, it achieves a balance between accuracy and efficiency. Experiments show that the recognition accuracy and detection speed of our method in the same environment have reached the level of mainstream algorithms, and an effective balance has been achieved between the two.

Download Full-text

Robust lane tracking algorithm for forward target detection of automated driving vehicles

Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering ◽

10.1177/0954407018785484 ◽

2018 ◽

Vol 233 (7) ◽

pp. 1930-1949 ◽

Cited By ~ 7

Author(s):

Moonhyung Song ◽

Changil Kim ◽

Moonsik Kim ◽

Kyongsu Yi

Keyword(s):

Conventional Method ◽

Lane Detection ◽

Model Parameters ◽

Detection Accuracy ◽

Vision Sensor ◽

Automated Driving ◽

Long Distance ◽

Virtual Sensor ◽

Lane Tracking ◽

Sensor Module

Detection of a primary forward target is one of the most important factors in a longitudinal control system of automated driving vehicles. If there is no special event such as a lane change, a lane detected by a vision sensor can be deemed as a predicted driving path of the ego vehicle. Based on this, it is possible to determine the state of the primary forward target vehicle accurately using the detected lanes. However, malfunction of vision sensors can induce unrecognized and misrecognized lane detection problems on account of internal/external environment factors. Furthermore, if a detection point is getting farther from the subject vehicle, the detection accuracy is getting lower because of inaccuracy of lane model parameters. To solve these problems, a novel method of lane tracking has been investigated. First, an integrated sensor module that combines a virtual sensor and a vision sensor has been developed. The virtual sensor is combined kinematic and dynamic vehicle model to be used in the full-speed range. Second, a lane estimator to improve lane detection accuracy at a long distance has been developed. The lane width in the public road can be assumed to be constant in the same road type. Based on this assumption, the clothoid parameters can be restored and consequently improve the lane detection accuracy. The performance of the proposed algorithm was verified by actual vehicle tests on public roads with manual driving and on proving ground with an automated driving system. The proposed algorithm has been compared with a conventional method which is based on in-vehicle yaw rate sensor. The test results have shown significant improvement of lane tracking performance over the conventional method.

Download Full-text

Purification of Urokinase from Complex Mixtures Using Immobilized Monoclonal Antibody Against Urokinase Light Chain

Thrombosis and Haemostasis ◽

10.1055/s-0038-1657308 ◽

1983 ◽

Vol 49 (01) ◽

pp. 024-027 ◽

Cited By ~ 20

Author(s):

David Vetterlein ◽

Gary J Calton

Keyword(s):

Monoclonal Antibody ◽

Light Chain ◽

Culture Media ◽

Biological Fluids ◽

Single Step ◽

High Yield ◽

Step Method ◽

Commercial Preparation ◽

E Coli ◽

Tissue Culture Media

SummaryThe preparation of a monoclonal antibody (MAB) against high molecular weight (HMW) urokinase light chain (20,000 Mr) is described. This MAB was immobilized and the resulting immunosorbent was used to isolate urokinase starting with an impure commercial preparation, fresh urine, spent tissue culture media, or E. coli broth without preliminary dialysis or concentration steps. Monospecific antibodies appear to provide a rapid single step method of purifying urokinase, in high yield, from a variety of biological fluids.

Download Full-text

THE PREDICTIVE CONTENT OF DISAGGREGATED NORMAL INCOME: An Empirical Study in the JSX

Gadjah Mada International Journal of Business ◽

10.22146/gamaijb.5633 ◽

2003 ◽

Vol 5 (3) ◽

pp. 363 ◽

Cited By ~ 1

Author(s):

Slamet Sugiri

Keyword(s):

Linear Regression ◽

Regression Models ◽

Stock Exchange ◽

Manufacturing Firms ◽

Single Step ◽

Model Parameters ◽

Linear Regression Models ◽

Operating Income ◽

Predictive Content ◽

Multiple Step

The main objective of this study is to examine a hypothesis that the predictive content of normal income disaggregated into operating income and nonoperating income outperforms that of aggregated normal income in predicting future cash flow. To test the hypothesis, linear regression models are developed. The model parameters are estimated based on fifty-five manufacturing firms listed in the Jakarta Stock Exchange (JSX) up to the end of 1997.This study finds that empirical evidence supports the hypothesis. This evidence supports arguments that, in reporting income from continuing operations, multiple-step approach is preferred to single-step one.

Download Full-text

The use of remote sensing satellite using deep learning in emergency monitoring of high-level landslides disaster in Jinsha River

The Journal of Supercomputing ◽

10.1007/s11227-020-03604-4 ◽

2021 ◽

Author(s):

Leijin Long ◽

Feng He ◽

Hongjiang Liu

Keyword(s):

Remote Sensing ◽

Southwest China ◽

Influence Factors ◽

Classification Error ◽

Model Parameters ◽

Detection Accuracy ◽

Remote Sensing Images ◽

Jinsha River ◽

Detection Model ◽

High Level

AbstractIn order to monitor the high-level landslides frequently occurring in Jinsha River area of Southwest China, and protect the lives and property safety of people in mountainous areas, the data of satellite remote sensing images are combined with various factors inducing landslides and transformed into landslide influence factors, which provides data basis for the establishment of landslide detection model. Then, based on the deep belief networks (DBN) and convolutional neural network (CNN) algorithm, two landslide detection models DBN and convolutional neural-deep belief network (CDN) are established to monitor the high-level landslide in Jinsha River. The influence of the model parameters on the landslide detection results is analyzed, and the accuracy of DBN and CDN models in dealing with actual landslide problems is compared. The results show that when the number of neurons in the DBN is 100, the overall error is the minimum, and when the number of learning layers is 3, the classification error is the minimum. The detection accuracy of DBN and CDN is 97.56% and 97.63%, respectively, which indicates that both DBN and CDN models are feasible in dealing with landslides from remote sensing images. This exploration provides a reference for the study of high-level landslide disasters in Jinsha River.

Download Full-text

On-CMOS Image Sensor Processing for Lane Detection

Sensors ◽

10.3390/s21113713 ◽

2021 ◽

Vol 21 (11) ◽

pp. 3713

Author(s):

Soyeon Lee ◽

Bohyeok Jeong ◽

Keunyeol Park ◽

Minkyu Song ◽

Soo Youn Kim

Keyword(s):

Edge Detection ◽

High Speed ◽

Cmos Image Sensor ◽

Image Sensor ◽

Image Resolution ◽

Lane Detection ◽

Frame Rate ◽

Total Power ◽

Detection Accuracy ◽

Total Power Consumption

This paper presents a CMOS image sensor (CIS) with built-in lane detection computing circuits for automotive applications. We propose on-CIS processing with an edge detection mask used in the readout circuit of the conventional CIS structure for high-speed lane detection. Furthermore, the edge detection mask can detect the edges of slanting lanes to improve accuracy. A prototype of the proposed CIS was fabricated using a 110 nm CIS process. It has an image resolution of 160 (H) × 120 (V) and a frame rate of 113, and it occupies an area of 5900 μm × 5240 μm. A comparison of its lane detection accuracy with that of existing edge detection algorithms shows that it achieves an acceptable accuracy. Moreover, the total power consumption of the proposed CIS is 9.7 mW at pixel, analog, and digital supply voltages of 3.3, 3.3, and 1.5 V, respectively.

Download Full-text

High-resolution limited-angle phase tomography of dense layered objects using deep neural networks

Proceedings of the National Academy of Sciences ◽

10.1073/pnas.1821378116 ◽

2019 ◽

Vol 116 (40) ◽

pp. 19848-19856 ◽

Cited By ~ 12

Author(s):

Alexandre Goy ◽

Girish Rughoobur ◽

Shuai Li ◽

Kwabena Arthur ◽

Akintunde I. Akinwande ◽

...

Keyword(s):

Integrated Circuit ◽

Deep Neural Networks ◽

Tomographic Reconstruction ◽

Synthetic Data ◽

Single Step ◽

Training Set ◽

Step Method ◽

3 Dimensional ◽

Dimensional Reconstruction ◽

Optical Domain

We present a machine learning-based method for tomographic reconstruction of dense layered objects, with range of projection angles limited to ±10○. Whereas previous approaches to phase tomography generally require 2 steps, first to retrieve phase projections from intensity projections and then to perform tomographic reconstruction on the retrieved phase projections, in our work a physics-informed preprocessor followed by a deep neural network (DNN) conduct the 3-dimensional reconstruction directly from the intensity projections. We demonstrate this single-step method experimentally in the visible optical domain on a scaled-up integrated circuit phantom. We show that even under conditions of highly attenuated photon fluxes a DNN trained only on synthetic data can be used to successfully reconstruct physical samples disjoint from the synthetic training set. Thus, the need for producing a large number of physical examples for training is ameliorated. The method is generally applicable to tomography with electromagnetic or other types of radiation at all bands.

Download Full-text