iOrigin的博客

During the past 17 years, technology has progressed at astronomical speeds. We have experienced not only the birth of many new technologies but also the miniaturization of both technologies new and old. It would have been audacious in the beginning of the new millennium to expect the amount of technological development. Perhaps the next biggest change to our lives is going to be self-driving or autonomous vehicles. Personally, my motivation was grounded in the fact that I received an opportunity to learn a tremendous amount of knowledge in both a niche, modern, highlytechnological field and gained experience in the different facets of software development such as software architecture, design and testing. Additionally, having full autonomy over the decision process and design choices added a great sense of responsibility. This thesis deals with lane detection and driving of such autonomous vehicles. The goal of this thesis is to program an existing autonomous car, built using a Raspberry Pi 2, an Arduino Mega and 4 DC motors, to successfully navigate a black and white mat which is supposed to function as a model of a racetrack. A small camera is mounted on top of the car and this is used in conjunction with different computer vision techniques to analyse data and correctly predict the course the car should navigate.

Connect your Mac to just about anything. Connect to routers, servers, firewalls, industrial control and IoT devices with ease. Telnet and Raw Socket Support NEW Serial now includes comprehensive support for Telnet, including the RFC2217 extension for remote serial port control supported by many serial device servers. Serial supports raw TCP sockets as well. Flawless Emulation Serial is a full-featured terminal emulator supporting Xterm, VT102, and ANSI terminal controls. This allows you to navigate the menu-driven interfaces found in many routers, firewalls and switches and use text based programs including emacs, vi, and nano as if you were connected over the network. No Drivers Serial includes built-in, reliable support for almost every serial device on the market, sparing you the hassle of finding, installing, and updating drivers.

Abstract: Lane detection is an important foundation in the development of intelligent vehicles. To address problems such as low detection accuracy of traditional methods and poor real-time performance of deep learning-based methodologies, a lane detection algorithm for intelligent vehicles in complex road conditions and dynamic environments was proposed. Firstly, converting the distorted image and using the superposition threshold algorithm for edge detection, an aerial view of the lane was obtained via region of interest extraction and inverse perspective transformation. Secondly, the random sample consensus algorithm was adopted to fit the curves of lane lines based on the third-order B-spline curve model, and fitting evaluation and curvature radius calculation were then carried out on the curve. Lastly, by using the road driving video under complex road conditions and the Tusimple dataset, simulation test experiments for lane detection algorithm were performed. The experimental results show that the average detection accuracy based on road driving video reached 98.49%, and the average processing time reached 21.5 ms. The average detection accuracy based on the Tusimple dataset reached 98.42%, and the average processing time reached 22.2 ms. Compared with traditional methods and deep learning-based methodologies, this lane detection algorithm had excellent accuracy and real-time performance, a high detection efficiency and a strong anti-interference ability. The accurate recognition rate and average processing time were significantly improved. The proposed algorithm is crucial in promoting the technological level of intelligent vehicle driving assistance and conducive to the further improvement of the driving safety of intelligent vehicles.

Training deep models for lane detection is challenging due to the very subtle and sparse supervisory signals in- herent in lane annotations. Without learning from much richer context, these models often fail in challenging sce- narios, e.g., severe occlusion, ambiguous lanes, and poor lighting conditions. In this paper, we present a novel knowl- edge distillation approach, i.e., Self Attention Distillation (SAD), which allows a model to learn from itself and gains substantial improvement without any additional supervision or labels. Specifically, we observe that attention maps ex- tracted from a model trained to a reasonable level would encode rich contextual information. The valuable contex- tual information can be used as a form of ‘free’ supervision for further representation learning through performing top- down and layer-wise attention distillation within the net- work itself. SAD can be easily incorporated in any feed- forward convolutional neural networks (CNN) and does not increase the inference time. We validate SAD on three pop- ular lane detection benchmarks (TuSimple, CULane and BDD100K) using lightweight models such as ENet, ResNet- 18 and ResNet-34. The lightest model, ENet-SAD, per- forms comparatively or even surpasses existing algorithms. Notably, ENet-SAD has 20 × fewer parameters and runs 10 × faster compared to the state-of-the-art SCNN [16], while still achieving compelling performance in all bench- marks. Our code is available at https://github. com/cardwing/Codes-for-Lane-Detection.

We introduce a network that directly predicts the 3D layout of lanes in a road scene from a single image. This work marks a first attempt to address this task with onboard sensing without assuming a known constant lane width or relying on pre-mapped environments. Our network architecture, 3D-LaneNet, applies two new concepts: intranetwork inverse-perspective mapping (IPM) and anchorbased lane representation. The intra-network IPM projection facilitates a dual-representation information flow in both regular image-view and top-view. An anchor-percolumn output representation enables our end-to-end approach which replaces common heuristics such as clustering and outlier rejection, casting lane estimation as an object detection problem. In addition, our approach explicitly handles complex situations such as lane merges and splits. Results are shown on two new 3D lane datasets, a synthetic and a real one. For comparison with existing methods, we test our approach on the image-only tuSimple lane detection benchmark, achieving performance competitive with stateof-the-art

Abstract—This paper describes the work that has been done in the design and development of a wireless OBD II fleet management system. The system aims to measure speed, distance and fuel consumption of vehicles for tracking and analysis purposes. An OBD II reader is designed to measure speed and mass air flow, from which distance and fuel consumption are also computed. This data is then transmitted via WiFi to a remote server. The system also implements GPS tracking to determine the location of the vehicle. A database management system is implemented at the remote server for the storage and management of transmitted data and a graphical user interface (GUI) is developed for analysing the transmitted data . Various qualification tests are conducted to verify the functionality of the system. The results demonstrate that the system is capable of reading the various parameters, and can successfully process, transmit and display the readings.

摘要: 针对目前在汽车后市场故障诊断仪诊断内容开发领域出现的由于平台复杂而导致诊断内容开发工程师效率低下的问题，对 开发平台结构的组成、ECU 模拟器软件的设计、J2534 协议接口的实现等方面进行了研究，提出了把调试平台简化为一台电脑，将 J2534 诊断接口协议应用到诊断软件和 ECU 模拟器的通信中的思路，实现了离线故障诊断的仿真; 开发了 ECU 模拟器及其与诊断 仪的通信接口，模拟器采用双进程模型和诊断仪进行了通信，其与已开发完成的故障诊断软件 DiagAnalyzer 间通过 J2534 协议所规定 的 API 完成了通信。研究和测试结果表明: 该 ECU 模拟器具备可行性和便捷性，诊断内容开发工程师只要在数据配置文件中写入特 定的诊断数据，即可完成对特定车型诊断仪数据的快速开发; 此外其可以和其他遵循 J2534 协议的故障诊断软件进行匹配和通信。 关键词: J2534 协议; ECU 模拟器; ISO15765; ＲPC 通信; 故障诊断仪

This specification provides the processor-specific definitions required by ELF [SCO-ELF10] for Arm based systems. The ELF specification is part of the larger System V ABI specification where it forms chapters 4 and 5. However, the specification can be used in isolation as a generic object and executable format. Platform Standards (page 14) of this document covers ELF related matters that are platform specific. Most of this material is related to the Base Platform ABI. Object Files (page 20) and Program Loading and Dynamic Linking (page 46) of this document are struc- tured to correspond to chapters 4 and 5 of the ELF specification. Specifically: • Object Files (page 20) covers object files and relocations • Program Loading and Dynamic Linking (page 46) covers program loading and dynamic linking. There are several drafts of the ELF specification on the SCO web site. This specification is based on the December 2003 draft, which was the most recent stable draft at the time this specification was developed.

USB的優點在於低廉易建置，缺點USB採用主從式架構(Master/slave system)，所有週邊裝置都要靠單一主控端 (PC)。 USB-IF 2001年12月公佈USB 2.0 On-The-Go (OTG)補充版 新增雙重角色裝置(Dual-role Device)及 HNP (Host Negotiation Protocol)、SRP (Session Request Protocol)兩種協定，使USB可以任意替換主控端(Host)和週邊端(Peripheral)的角色

Classical scheduling problems revolve around setting the length of the scheduling quantum to provide interactive re- sponsiveness while minimizing the context switch overhead, simultaneously catering to batch and interactive workloads in a single system, and efficiently managing scheduler run queues. By and large, by the year 2000, operating systems designers considered scheduling to be a solved problem; the Linus Torvalds quote is an accurate reflection of the general opinion at that time.

最佳的比较工具 Araxis Merge是一个可视化的文件比较、合并和同步的软件。能够进行版本和质量控制，创建 HTML 或是 XML 报告。能够方便的被软件工程师和 web 站点开发者使用快速精确地比较、了解和联合不同版本的源文件；进行版本和质量控制，创建 HTML 或是 XML 报告。 [1] Araxis Merge获得了美国IT杂志Dr. Dobb's颁发的Jolt Awards 2013 Jolt生产力奖。 [2]

What follows is a set of rules, guidelines, and tips that we have found to be useful in making C++ code portable across many machines and compilers. This information is the result of porting large amounts of code across about 25 different machines, and at least a dozen different C++ compilers. Some of these things will frustrate you and make you want to throw your hands up and say, ``well, that's just a stupid compiler if it doesn't do <insert favorite C++ feature>.'' But this is the reality of portable code. If you play by the rules, your code will seamlessly work on all of the Mozilla platforms and will be easy to port to newer machines.

This document and its associated software may be used as stated in the underlying license agreement. Rational Software Corporation expressly disclaims all other warranties, express or implied, with respect to the media and software product and its documentation, including without limitation, the warranties of merchantability or fitness for a particular purpose or arising from a course of dealing, usage, or trade practice.

Programming is largely a manual effort, which is tedious and time consuming. • Developers often wonder what code to write in order to complete a programming task • Unfamiliar problems • Unfamiliar APIs •... • In a survey conducted at Microsoft, 67.6% respondents mentioned that they encountered problems in learning APIs.

什么是团队全栈? 为什么要团队全栈? 技术团队遇到的问题

C++反射的应用与实践 程序对自身程序结构认知的能力 • 能够使用数据结构来描述数据类型和其成员 • 能够主动地遍历这些成员

RK3288 开发板上默认 QT 版本为 5.5.1，为保证良好兼容性，因此选择 Q5.5.1. 下载链接: http://download.qt.io/archive/qt/5.5/5.5.1/single/qt-everywhere-opensource-src-5.5.1. tar.xz 建议:代码解压后最好在本地用 git 管理起来，这样方便还原以及清理。想重新编译只需要 git clean -df 就可以还原，非常方便。

有限群是具有有限多个元素的群。群论的重要内容之一。其所含元素的个数，称为有限群的阶。有限群可分为两大类：可解群与非可解群（特别包括非交换单群）（见群、有限单群）。

本文档主要对蓝牙电话整体结构进行分析，包括主体架构、协议栈、主体流程、以及协议说明。为下一 步的开发提供帮助