WO2019114731A1 - 自动泊车的方法和设备、智能汽车以及计算机存储介质 - Google Patents

自动泊车的方法和设备、智能汽车以及计算机存储介质 Download PDF

Info

Publication number
WO2019114731A1
WO2019114731A1 PCT/CN2018/120473 CN2018120473W WO2019114731A1 WO 2019114731 A1 WO2019114731 A1 WO 2019114731A1 CN 2018120473 W CN2018120473 W CN 2018120473W WO 2019114731 A1 WO2019114731 A1 WO 2019114731A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
limit
video
identifier
real
Prior art date
Application number
PCT/CN2018/120473
Other languages
English (en)
French (fr)
Inventor
王凯
周煜远
丁晨曦
李谦
Original Assignee
蔚来汽车有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 蔚来汽车有限公司 filed Critical 蔚来汽车有限公司
Priority to EP18888380.5A priority Critical patent/EP3725626B1/en
Publication of WO2019114731A1 publication Critical patent/WO2019114731A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D15/00—Steering not otherwise provided for
    • B62D15/02—Steering position indicators ; Steering position determination; Steering aids
    • B62D15/027—Parking aids, e.g. instruction means
    • B62D15/0285—Parking performed automatically
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/06—Automatic manoeuvring for parking
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D15/00—Steering not otherwise provided for
    • B62D15/02—Steering position indicators ; Steering position determination; Steering aids
    • B62D15/027—Parking aids, e.g. instruction means
    • B62D15/028—Guided parking by providing commands to the driver, e.g. acoustically or optically
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2420/00—Indexing codes relating to the type of sensors based on the principle of their operation
    • B60W2420/40—Photo, light or radio wave sensitive means, e.g. infrared sensors
    • B60W2420/403—Image sensing, e.g. optical camera
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2552/00—Input parameters relating to infrastructure
    • B60W2552/53—Road markings, e.g. lane marker or crosswalk

Definitions

  • the present invention relates to the field of automotive technology, and more particularly to a method and apparatus for automatic parking, a smart car, and a computer storage medium.
  • the battery-based power-exchange mode combined with large-scale centralized charging has become a competitive commercial technology model for the development of electric vehicles.
  • the power-changing mode is to replace the battery of the vehicle that is about to be exhausted or exhausted with a fully charged battery. In mechanical operation, it only takes a few minutes to complete the battery replacement. Compared with the charging mode, it reduces the charging time. The trouble is coming.
  • Chinese patent CN106004515A discloses a vehicle speed control method and system for automatic parking of electric vehicles.
  • the system includes a vehicle speed sensor module, an ultrasonic sensor, a vehicle positioning module, a parking controller and a vehicle speed controller.
  • the vehicle speed sensor module is used for detecting the moving speed of the electric vehicle
  • the ultrasonic sensor is used for collecting obstacle information around the electric vehicle, and the vehicle is positioned.
  • the module is used for acquiring the current position information of the electric vehicle
  • the parking controller is configured to set the ideal parking speed according to the collected obstacle information, the current position information and the preset parking trajectory, and the vehicle speed controller is used according to the moving speed and the ideal speed.
  • the parking speed adjusts the motor torque of the electric vehicle.
  • the vehicle positioning accuracy error of the ultrasonic sensor is generally 10-20 cm, and the positioning accuracy does not meet the power-requiring requirement.
  • the present invention provides a method of automatic parking, the method comprising: identifying that a vehicle is in a predetermined area, the predetermined area being defined by a plurality of limit signs Using a video capture device installed on both sides of the vehicle to perform video extraction on the environment surrounding the vehicle in real time; using a color space to locate a first limit mark in the video; and through morphological detection, according to the first a feature of the limit mark, the real-time position of the vehicle is derived; and the vehicle is guided to the target area according to the real-time position of the vehicle.
  • the predetermined area is a rectangular parking space, and the four corners of the parking space are respectively defined by corresponding L-shaped limit identifiers.
  • the plurality of limit identifiers have a color feature that is different from an environment of the periphery of the vehicle, and the use of the color space to locate the first limit identifier in the video includes: The first limit identifier is located in the video.
  • the morphological detection, according to the feature of the first limit identifier, obtaining the real-time location of the vehicle includes: determining, by using a morphological detection, a feature of the first limit identifier; Determining the absolute position of the first limit identifier; determining, by the pixel of the video, and estimating the absolute position of the first limit identifier, the video capture device is relatively a distance to the target area; and determining a real-time location of the vehicle based at least on a distance of the video capture device relative to the target area.
  • the video capture device is mounted on a rear view mirror of the vehicle, and the visible range thereof covers at least one of the plurality of limit identifiers.
  • an apparatus for automatic parking comprising: identification means for identifying that a vehicle is in a predetermined area, the predetermined area being defined by a plurality of limit marks; and a video capture device installed On both sides of the vehicle, for real-time video extraction of the environment around the vehicle; positioning means for locating the first limit identifier in the video by using a color space; and detecting means for adopting morphology Detecting, according to the feature of the first limit identifier, obtaining a real-time location of the vehicle; and guiding means for guiding the vehicle to the target area according to the real-time position of the vehicle.
  • the predetermined area is a rectangular parking space, and the four corners of the parking space are respectively defined by corresponding L-shaped limit identifiers.
  • the plurality of limit identifiers have color features different from an environment of the periphery of the vehicle, and the positioning device is configured to locate the first limit from the video by color patch analysis logo.
  • the detecting device includes: a first determining unit, configured to determine a feature of the first limit identifier by morphological detection; and a second determining unit, configured to be used according to the first limit identifier Determining an absolute position of the first limit identifier; determining a unit for estimating the video capture device by comparing the pixels of the video and determining an absolute position of the first limit identifier a distance of the target area; and a third determining unit configured to determine a real-time location of the vehicle based at least on a distance of the video capture device relative to the target area.
  • the video capture device is mounted on a rear view mirror of the vehicle, and the visible range thereof covers at least one of the plurality of limit identifiers.
  • a smart car comprising an automatic parking device as described above.
  • a computer storage medium includes instructions that, when executed, perform the aforementioned method of automatic parking.
  • the technical solution of the invention can provide high-precision positioning for new energy vehicles when they are automatically parked at the power station.
  • the traditional positioning method mainly relies on ultrasonic technology, which has high dependence on the surrounding physical materials and shapes, and the positioning accuracy is also low.
  • the technical solution of the present invention applies the technology of video positioning to the positioning of the vehicle position, sets a parking space at a designated position of the power-changing area, and uses the video capture device (such as a camera) of the vehicle body to extract the video and extract the parking space feature.
  • High-precision positioning for vehicle position, the final positioning error is only 1 ⁇ 2cm.
  • the present invention can utilize the L-shaped mark that is common when the parking space is currently set as the key position information of the detected video, thereby positioning the position of the vehicle during the automatic parking. Since the L-shaped sign is generally set when the parking space is demarcated, the solution of the present invention can be applied to various practical application scenarios, including but not limited to parking lot parking, community parking, power-changing parking, etc. without additional Set other location identifiers.
  • FIG. 1 is a schematic view showing an automatic parking method according to an embodiment of the present invention
  • Figure 2 is a block diagram showing the structure of an automatic parking apparatus according to an embodiment of the present invention.
  • Figure 3 is a plan view showing a limit mark of an embodiment of the present invention.
  • Fig. 4 is a schematic view showing the scene of automatic parking according to an embodiment of the present invention.
  • morphological detection refers to the detection of the shape (including direction, shape, etc.) of a target object.
  • performing a morphological detection of the limit identification indicates extracting a shape feature of the limit identification.
  • the automatic parking method 1000 includes the following steps:
  • step 110 identifying that the vehicle is in a predetermined area, the predetermined area being defined by a plurality of limit identifiers;
  • step 130 video extraction is performed on the environment surrounding the vehicle in real time by using a video capture device installed on both sides of the vehicle;
  • step 150 the first limit identifier in the video is located by using a color space
  • step 170 the real-time location of the vehicle is obtained according to the characteristics of the first limit identifier by morphological detection
  • step 190 the vehicle is directed to the target area based on the real-time location of the vehicle.
  • the video localization method described above is based on color space detection and morphological detection to locate a vehicle in a predetermined area so that it can be automatically parked into a target area, such as a docking station for automatic power exchange. Based on video positioning, the positioning accuracy can be greatly improved, so as to meet the requirements of the vehicle for automatic parking in the substation for positioning accuracy.
  • the predetermined area is a rectangular parking space, and the four corners of the parking space are respectively defined by corresponding L-shaped limit identifiers.
  • the plurality of limit markers have color features that are distinct from an environment of the perimeter of the vehicle. In this way, it is more advantageous to identify the limit mark from the video and improve the accuracy of the recognition.
  • step 150 may locate the first limit identifier from the video by patch analysis.
  • the step 170 includes: determining, by using a morphological detection, a feature of the first limit identifier; determining an absolute position of the first limit identifier according to the notch feature of the first limit identifier Deriving a distance of the video capture device relative to the target area by calibrating pixels of the video and based on an absolute position of the first limit mark; and at least depending on a distance of the video capture device relative to the target area, A real-time location of the vehicle is determined.
  • the vehicle may obtain map information of the substation from the cloud, the map information including the absolute location of each of the limit identifiers.
  • the first limit identifier in the video is determined from the plurality of limit identifiers, and the absolute position of the first limit identifier is determined according to the map information.
  • the relative position change can be performed, that is, according to the detected absolute position of the mark and the position of the body relative to the body, the absolute position of the current vehicle and the target area are obtained (for example, The relative position of the tram position, thus giving the vehicle a specific location in the parking process in real time.
  • the video capture device is mounted on a rear view mirror of the vehicle, the visual extent of which covers at least one of the plurality of limit markers.
  • the video capture device (not shown) is mounted on the rear view mirror of the vehicle 400 and its field of view generally covers a semi-circular area, it is located on the left.
  • the video capture device on the side can capture the limit markers 402 and 406 on the left side.
  • the video capture device on the right can capture the limit markers 404 and 408 on the right.
  • the limit markers 402-408 define the shape of the predetermined area 410.
  • the vehicle 400 is parked in the lower half of the predetermined area 410, since the video capture device (not shown) is mounted on the rear view mirror of the vehicle 400, the video capture devices on both sides of the vehicle only The limit identifiers 406 and 408 located at the left rear and the right rear of the predetermined area can be separately collected, that is, the video capture device on one side can only collect the video of one limit mark.
  • the video capture devices on both sides of the vehicle can only collect the limit signs 402 and 404 located in the left front and right front of the predetermined area, respectively. Finally, the vehicle 400 is automatically docked to the target area 420.
  • FIG. 2 provides an apparatus 2000 for automatic parking.
  • the device 2000 includes an identification device 210, a video capture device 220, a positioning device 230, a detection device 240, and a guiding device 250.
  • the identification device 210 is configured to identify that the vehicle is in a predetermined area, and the predetermined area is defined by a plurality of limit identifiers.
  • a video capture device 220 is mounted on both sides of the vehicle for video extraction of the environment surrounding the vehicle in real time.
  • the positioning device 230 is configured to use a color space to locate a first limit identifier in the video.
  • the detecting device 240 is configured to obtain a real-time location of the vehicle according to the feature of the first limit identifier by morphological detection.
  • the guiding device 250 is for guiding the vehicle to the target area according to the real-time position of the vehicle.
  • the predetermined area is a rectangular parking space, and the four corners of the parking space are each defined by a corresponding L-shaped limit identification.
  • the plurality of limit markers have color features that are distinct from an environment of the perimeter of the vehicle. In this way, it is more advantageous to identify the limit mark from the video and improve the accuracy of the recognition.
  • the positioning device 230 can be further configured to locate the first limit identifier from the video by patch analysis.
  • the detecting device 240 may include: a first determining unit, configured to determine a feature of the first limit identifier by morphological detection; and a second determining unit, configured to use the first limit identifier Determining the absolute position of the first limit identifier; the estimating unit is configured to estimate the relative position of the video capture device by calibrating the pixels of the video and according to the absolute position of the first limit identifier a distance to the target area; and a third determining unit configured to determine a real-time location of the vehicle based at least on a distance of the video capture device relative to the target area.
  • the video capture device 220 is mounted on a rear view mirror of the vehicle, the visual extent of which covers at least one of the plurality of limit markers.
  • the foregoing method embodiments of the present invention can be implemented as a computer program product.
  • the invention may be embodied, for example, in the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) including computer usable program code.
  • Figure 3 shows a design of a guide mark for vehicle position detection.
  • the identifier is a limit identifier of a predetermined area. For example, it may be considered to arrange it, for example, at a designated location of the substation (if there is no such limit) to define the parking area of the vehicle that is about to be replaced.
  • the video is captured by using a camera on the rearview mirror of the vehicle, and the color space detection and morphological detection methods are used to detect the parking space features in the video.
  • the area 410 is a vehicle parking start area (ie, a predetermined area) defined by the limit signs 402 to 408 on the four corners, and the area 420 is a power station parking space (ie, the target area). ).
  • the limit mark will play a role in two stages. The first is the initial positioning area. When the power is changed, the vehicle will be parked in the parking space area. The identification system will automatically identify the initial position of the vehicle to give parking start information. Second, the parking process In the figure, as shown in Figure 4, the identification system will identify the vehicle parking trajectory and assist the vehicle to safely enter the power station.
  • the solution of the present invention mainly uses video acquisition devices installed on both sides of the vehicle to perform video extraction on the surrounding environment, and detects important position information in the video through the method of color space detection and morphology detection, including the limit identification and target.
  • the position of the area, etc. is calculated based on the spatial transformation of these positions, and the position of the vehicle relative to the target area is further calculated, thereby locating the position of the vehicle during the automatic parking.
  • the technical solution of the present invention applies the video positioning technology to the positioning of the vehicle position, sets the parking space at the designated position of the power-changing area, and uses the video capturing device of the vehicle body (for example, the camera).
  • the video is extracted and the parking space features are extracted for high-precision positioning of the vehicle position, and the final positioning error is only 1 to 2 cm.

Landscapes

  • Engineering & Computer Science (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Automation & Control Theory (AREA)
  • Image Analysis (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种自动泊车的方法和设备、智能汽车以及计算机存储介质，其中方法包括：识别车辆（400）处于预定区域（410），预定区域（410）由多个限位标识（402、404、406、408）所限定；利用安装在车辆两侧的视频采集装置（220）实时对车辆（400）周边的环境进行视频提取；运用颜色空间定位出视频中的第一限位标识；通过形态学检测，根据第一限位标识的特征，得出车辆（400）的实时位置；以及根据车辆（400）的实时位置，将车辆导引至目标区域（420）。

Description

自动泊车的方法和设备、智能汽车以及计算机存储介质 技术领域
本发明涉及汽车技术领域，特别涉及自动泊车的方法和设备、智能汽车以及计算机存储介质。
背景技术
电动汽车作为一种新能源交通工具，具有噪音低、能源利用效率高、无移动废弃排放等特点，已成为我国重点支持的战略性新兴产业之一。
基于电池租赁的换电模式配合大规模集中型充电已经成为当前电动汽车发展具有竞争力的商业技术模式。换电模式就是用充满电的电池更换下即将或已经耗尽的车辆电池，在机械操作下，它只需要几分钟就能完成电池的更换，相比于充电模式，它减少了充电时间长带来的麻烦。
目前国内已建立有多座充换电站，为电动汽车的动力电池提供充换电服务。但由于充换电站对车辆停车位置的要求较高，目前主要依赖超声波技术的车辆定位精度难以满足实际需要。
中国专利CN106004515A公开了一种用于电动汽车自动泊车的车速控制方法及系统。该系统包括车速传感器模块、超声波传感器、车辆定位模块、泊车控制器和车速控制器，车速传感器模块用于检测电动汽车的运动车速，超声波传感器用于采集电动汽车周围的障碍物信息，车辆定位模块用于采集获得电动汽车的当前位置信息，泊车控制器用于根据采集的障碍物信息、当前位置信息以及预设的泊车轨迹设定理想泊车车速，车速控制器用于根据运动车速和理想泊车车速调节电动汽车的电机扭矩。
上述专利尽管可在低速情况下实现误差较小的泊车速度控制功能，但超声波传感器的车辆定位精度误差一般在10-20cm，该定位精度达不到换电要求。
以上公开于本发明背景部分的信息仅仅旨在增加对本发明的总 体背景的理解，而不应当被视为承认或以任何形式暗示该信息构成已为本领域一般技术人员所公知的现有技术。
发明内容
为了解决以上现有技术的多个问题中的至少一个，本发明提供了一种自动泊车的方法，所述方法包括：识别车辆处于预定区域，所述预定区域由多个限位标识所限定；利用安装在所述车辆两侧的视频采集装置实时对所述车辆周边的环境进行视频提取；运用颜色空间定位出所述视频中的第一限位标识；通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置；以及根据所述车辆的实时位置，将所述车辆导引至目标区域。
在上述方法中，所述预定区域为矩形的停车位，所述停车位的四个角分别由对应的L形限位标识所限定。
在上述方法中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征，并且运用颜色空间定位出所述视频中的第一限位标识包括：通过色块分析，从所述视频中定位出所述第一限位标识。
在上述方法中，通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置包括：通过形态学检测，确定所述第一限位标识的特征；根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。
在上述方法中，所述视频采集装置安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。
根据本发明的又一个方面，提供了一种自动泊车的设备，包括：识别装置，用于识别车辆处于预定区域，所述预定区域由多个限位标识所限定；视频采集装置，其安装在所述车辆两侧，用于实时对所述车辆周边的环境进行视频提取；定位装置，用于运用颜色空间定位出所述视频中的第一限位标识；检测装置，用于通过形态学检测，根据 所述第一限位标识的特征，得出所述车辆的实时位置；以及导引装置，用于根据所述车辆的实时位置，将所述车辆导引至目标区域。
在上述设备中，所述预定区域为矩形的停车位，所述停车位的四个角分别由对应的L形限位标识所限定。
在上述设备中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征，并且所述定位装置配置成通过色块分析，从所述视频中定位出所述第一限位标识。
在上述设备中，所述检测装置包括：第一确定单元，用于通过形态学检测，确定所述第一限位标识的特征；第二确定单元，用于根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；推定单元，用于通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及第三确定单元，用于至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。
在上述设备中，所述视频采集装置安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。
根据本发明的又一个方面，提供了一种智能汽车，其包括如前所述的自动泊车的设备。
根据本发明的又一个方面，提供了一种计算机存储介质，其包括指令，所述指令在被运行时执行前述自动泊车的方法。
本发明的技术方案能够为新能源车辆在换电站自动泊车时提供高精度定位。传统的定位方法主要依赖超声波技术，对周边的物理材质和形状依赖度较高，另外定位精度也较低。本发明的技术方案将视频定位的技术运用到车辆位置的定位上，在换电区域指定位置设置停车位，运用车辆周身的视频采集装置(例如摄像头)提取视频并对停车位特征进行提取，以用于车辆位置的高精度定位，最终定位的误差仅为1～2cm。
此外，本发明可利用当前设置停车位时常见的L形标识来作为检测视频的关键位置信息，从而在自动泊车的过程中对车辆所处位置进行定位。由于这种L形标识一般在划定停车位时都会设置，所以本发 明的方案可适用于各种实际应用场景，包括但不限于停车场停车、小区停车、换电站停车等等，而无需额外地设置其他定位标识。
通过纳入本文的附图以及随后与附图一起用于说明本发明的某些原理的具体实施方式，本发明的方法和装置所具有的其它特征和优点将更为具体地变得清楚或得以阐明。
附图说明
图1是表示本发明的一个实施例的自动泊车方法的示意图；
图2是表示本发明的一个实施例的自动泊车设备的结构示意图；
图3是表示本发明的一个实施例的限位标识的设计图；以及
图4是表示本发明的一个实施例的自动泊车的场景示意图。
具体实施方式
以下说明描述了本发明的特定实施方式以教导本领域技术人员如何制造和使用本发明的最佳模式。为了教导发明原理，已简化或省略了一些常规方面。本领域技术人员应该理解源自这些实施方式的变型将落在本发明的范围内。本领域技术人员应该理解下述特征能够以各种方式接合以形成本发明的多个变型。由此，本发明并不局限于下述特定实施方式，而仅由权利要求和它们的等同物限定。
在本说明书中，诸如“包含”和“包括”之类的用语表示除了具有在说明书和权利要求书中有直接和明确表述的单元和步骤以外，本发明的技术方案也不排除具有未被直接或明确表述的其它单元和步骤的情形。
诸如“第一”和“第二”之类的用语并不表示单元在时间、空间、大小等方面的顺序而仅仅是作区分各单元之用。
在本说明的上下文中，“形态学检测”指对目标物体的形态(包括方向、形状等)进行检测。在一个实施例中，对限位标识进行形态学检测表示提取该限位标识的形状特征。
图1是表示本发明的一个实施例的自动泊车方法1000的示意图。其中，自动泊车方法1000包括如下步骤：
在步骤110中，识别车辆处于预定区域，所述预定区域由多个限位标识所限定；
在步骤130中，利用安装在所述车辆两侧的视频采集装置实时对所述车辆周边的环境进行视频提取；
在步骤150中，运用颜色空间定位出所述视频中的第一限位标识；
在步骤170中，通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置；以及
在步骤190中，根据所述车辆的实时位置，将所述车辆导引至目标区域。
上述视频定位方法基于颜色空间检测和形态学检测，对于处于预定区域的车辆进行定位，从而使其能够自动泊入目标区域，例如泊入换电站以进行自动换电。基于视频定位可以大幅度提高定位精度，从而满足车辆在换电站自动泊车对于定位精度的要求。
在一个实施例中，所述预定区域为矩形的停车位，所述停车位的四个角分别由对应的L形限位标识所限定。通过利用当前设置停车位时常见的L形标识来作为检测视频的关键位置信息，本发明的方案可适用于各种实际应用场景，包括但不限于停车场停车、小区停车、换电站停车等等，而无需额外地设置其他定位标识。
在一个实施例中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征。这样，可更有利于从视频中识别出限位标识，提高识别的准确度。在这个实施例中，步骤150可通过色块分析，从所述视频中定位出所述第一限位标识。
在一个实施例中，步骤170包括：通过形态学检测，确定所述第一限位标识的特征；根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。在一个实施例中，车辆可从云端获取换电站的地图信息，该地图信息包括各个限位标识的绝对位置。通过颜色空间检测和形态学检测，从多个限位标识中确定出视频中的第一限位标识，并根 据地图信息确定出第一限位标识的绝对位置。在确定出第一限位标识的绝对位置后，可进行相对位置变换，即根据所检测的标识的绝对位置以及其相对车身的位置，得出当前车辆的绝对位置以及相对于目标区域(例如换电车位)的相对位置，从而实时给出车辆在泊车过程中的具体位置。
在一个实施例中，所述视频采集装置安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。参考图4，如果车辆400停在预定区域410的中部时，由于视频采集装置(未示出)安装在车辆400的后视镜上且其视野一般能覆盖一个半圆形的区域，所以位于左侧的视频采集装置能采集到位于左侧的限位标识402以及406。类似地，位于右侧的视频采集装置能采集到位于右侧的限位标识404以及408。限位标识402-408限定出预定区域410的形状。在另一个实施例中，如果车辆400停在预定区域410的下半部时，由于视频采集装置(未示出)安装在车辆400的后视镜上，在车辆的两侧的视频采集装置只能分别采集到位于预定区域左后方以及右后方的限位标识406和408，即一侧的视频采集装置只能采集到一个限位标识的视频。在又一个实施例，如果车辆400停在预定区域410的上半部时，在车辆的两侧的视频采集装置只能分别采集到位于预定区域左前方以及右前方的限位标识402和404。最终，车辆400自动泊入到目标区域420。
图2提供了一种自动泊车的设备2000。该设备2000包括：识别装置210、视频采集装置220、定位装置230、检测装置240以及导引装置250。其中，识别装置210用于识别车辆处于预定区域，所述预定区域由多个限位标识所限定。视频采集装置220安装在所述车辆两侧，其用于实时对所述车辆周边的环境进行视频提取。定位装置230用于运用颜色空间定位出所述视频中的第一限位标识。检测装置240用于通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置。导引装置250用于根据所述车辆的实时位置，将所述车辆导引至目标区域。
在一个实施例中，所述预定区域为矩形的停车位，所述停车位的 四个角分别由对应的L形限位标识所限定。通过利用当前设置停车位时常见的L形标识来作为检测视频的关键位置信息，本发明的方案可适用于各种实际应用场景，包括但不限于停车场停车、小区停车、换电站停车等等，而无需额外地设置其他定位标识。
在一个实施例中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征。这样，可更有利于从视频中识别出限位标识，提高识别的准确度。在这个实施例中，定位装置230可进一步配置成通过色块分析，从所述视频中定位出所述第一限位标识。
在一个实施例中，检测装置240可包括：第一确定单元，用于通过形态学检测，确定所述第一限位标识的特征；第二确定单元，用于根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；推定单元，用于通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及第三确定单元，用于至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。
在一个实施例中，所述视频采集装置220安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。
本领域内的技术人员应明白，本发明的前述方法实施例可实现为计算机程序产品。因此，本发明例如可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
图3展示了用于车辆位置检测的导向标识的设计图。如图3所示，该标识为一预定区域的限位标识。例如，可考虑将其布置于例如换电站指定的位置(若无该限位标识)，用以限定即将进行换电的车辆的停车区域。在一个实施例中，通过运用车辆后视镜上的摄像头采集视频，运用颜色空间检测和形态学检测的方法，对视频中的停车位特征进行检测。
如图4所示，区域410为车辆泊车起始区域(即预定区域)，该区域由四个角落上的限位标识402至408来进行限定，区域420为换电站停车位(即目标区域)。限位标识将在两个阶段起到作用，一是 初始定位区域，换电时将车辆停在停车位区域内，识别系统将自动识别车辆初始位置给出泊车启动信息；二是泊车过程中，如图4所示，识别系统将识别车辆泊车轨迹，协助车辆安全进入换电站。
综上，本发明的方案主要运用车辆两侧安装的视频采集装置对周边的环境进行视频提取，通过颜色空间检测额和形态学检测的方法检测视频中重要的位置信息，包括限位标识、目标区域的位置等，根据这些位置进行空间变换计算，进一步计算出车辆相对于目标区域的位置，从而在自动泊车的过程中对车辆所处位置进行定位。
与传统主要依赖超声波技术的定位方法相比，本发明的技术方案将视频定位的技术运用到车辆位置的定位上，在换电区域指定位置设置停车位，运用车辆周身的视频采集装置(例如摄像头)提取视频并对停车位特征进行提取，以用于车辆位置的高精度定位，最终定位的误差仅为1～2cm。
以上例子主要说明了本发明的自动泊车的方法和设备、智能汽车以及计算机存储介质。尽管只对其中一些本发明的一个或多个实施例进行了描述，但是本领域普通技术人员应当了解，本发明可以在不偏离其主旨与范围内以许多其他的形式实施。因此，所展示的例子与实施方式被视为示意性的而非限制性的，在不脱离如所附各权利要求所定义的本发明精神及范围的情况下，本发明可能涵盖各种的修改与替换。

Claims (12)

  1. 一种自动泊车的方法，所述方法包括：
    识别车辆处于预定区域，所述预定区域由多个限位标识所限定；
    利用安装在所述车辆两侧的视频采集装置实时对所述车辆周边的环境进行视频提取；
    运用颜色空间定位出所述视频中的第一限位标识；
    通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置；以及
    根据所述车辆的实时位置，将所述车辆导引至目标区域。
  2. 如权利要求1所述的方法，其中，所述预定区域为矩形的停车位，所述停车位的四个角分别由对应的L形限位标识所限定。
  3. 如权利要求1或2所述的方法，其中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征，并且运用颜色空间定位出所述视频中的第一限位标识包括：
    通过色块分析，从所述视频中定位出所述第一限位标识。
  4. 如权利要求1或2所述的方法，其中，通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置包括：
    通过形态学检测，确定所述第一限位标识的特征；
    根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；
    通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及
    至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。
  5. 如权利要求1所述的方法，其中，所述视频采集装置安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。
  6. 一种自动泊车的设备，包括：
    识别装置，用于识别车辆处于预定区域，所述预定区域由多个 限位标识所限定；
    视频采集装置，其安装在所述车辆两侧，用于实时对所述车辆周边的环境进行视频提取；
    定位装置，用于运用颜色空间定位出所述视频中的第一限位标识；
    检测装置，用于通过形态学检测，根据所述第一限位标识的特征，得出所述车辆的实时位置；以及
    导引装置，用于根据所述车辆的实时位置，将所述车辆导引至目标区域。
  7. 如权利要求6所述的设备，其中，所述预定区域为矩形的停车位，所述停车位的四个角分别由对应的L形限位标识所限定。
  8. 如权利要求6或7所述的设备，其中，所述多个限位标识具有区别于所述车辆周边的环境的颜色特征，并且所述定位装置配置成通过色块分析，从所述视频中定位出所述第一限位标识。
  9. 如权利要求6或7所述的设备，其中，所述检测装置包括：
    第一确定单元，用于通过形态学检测，确定所述第一限位标识的特征；
    第二确定单元，用于根据所述第一限位标识的缺角特征，确定所述第一限位标识的绝对位置；
    推定单元，用于通过对所述视频的像素标定并根据所述第一限位标识的绝对位置，推定所述视频采集装置相对于目标区域的距离；以及
    第三确定单元，用于至少根据所述视频采集装置相对于目标区域的距离，确定所述车辆的实时位置。
  10. 如权利要求6所述的设备，其中，所述视频采集装置安装在所述车辆的后视镜上，其可视范围覆盖多个限位标识中的至少一个限位标识。
  11. 一种智能汽车，其包括如权利要求6-10中任一项所述的自动泊车的设备。
  12. 一种计算机存储介质，其包括指令，所述指令在被运行时 执行如权利要求1至5中任一项所述的自动泊车的方法。
PCT/CN2018/120473 2017-12-15 2018-12-12 自动泊车的方法和设备、智能汽车以及计算机存储介质 WO2019114731A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP18888380.5A EP3725626B1 (en) 2017-12-15 2018-12-12 Automatic parking method and device, intelligent vehicle and computer storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711346034.6 2017-12-15
CN201711346034.6A CN108569280B (zh) 2017-12-15 2017-12-15 自动泊车的方法和设备、智能汽车以及计算机存储介质

Publications (1)

Publication Number Publication Date
WO2019114731A1 true WO2019114731A1 (zh) 2019-06-20

Family

ID=63576670

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/120473 WO2019114731A1 (zh) 2017-12-15 2018-12-12 自动泊车的方法和设备、智能汽车以及计算机存储介质

Country Status (3)

Country Link
EP (1) EP3725626B1 (zh)
CN (1) CN108569280B (zh)
WO (1) WO2019114731A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108569280B (zh) * 2017-12-15 2020-03-24 上海蔚来汽车有限公司 自动泊车的方法和设备、智能汽车以及计算机存储介质
CN111619552A (zh) * 2020-05-26 2020-09-04 三一专用汽车有限责任公司 适于施工区域的工程车辆自主泊车系统和自主泊车方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102933442A (zh) * 2010-06-09 2013-02-13 法雷奥开关和传感器有限责任公司 用于当在停车位中停放时辅助机动车辆的驾驶员的方法、驾驶员辅助装置和机动车辆
CN202923500U (zh) * 2012-11-27 2013-05-08 德尔福电子(苏州)有限公司 电动车辆无线充电引导系统
CN106004515A (zh) 2016-05-12 2016-10-12 广州橙行智动汽车科技有限公司 用于电动汽车自动泊车的车速控制方法及系统
DE102016013180A1 (de) * 2016-11-04 2017-05-24 Daimler Ag Verfahren zum Positionieren eines Kraftfahrzeugs relativ zu einer induktiven Ladeeinrichtung
CN108569280A (zh) * 2017-12-15 2018-09-25 蔚来汽车有限公司 自动泊车的方法和设备、智能汽车以及计算机存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5088074B2 (ja) * 2007-10-01 2012-12-05 日産自動車株式会社 駐車支援装置及び方法
JP4530060B2 (ja) * 2008-02-19 2010-08-25 トヨタ自動車株式会社 駐車支援装置及び方法
JP2013018406A (ja) * 2011-07-12 2013-01-31 Fujitsu Ten Ltd 駐車枠の認識装置、駐車枠の認識方法、及び、プログラム
DE102011087797A1 (de) * 2011-12-06 2013-06-06 Robert Bosch Gmbh Verfahren und Vorrichtung zur Lokalisation einer vordefinierten Parkposition
CN104943686B (zh) * 2014-03-28 2018-11-06 北京奇虎科技有限公司 基于无线信号识别的自动泊车实现方法和装置
CN104260722B (zh) * 2014-09-23 2017-06-06 北京理工大学 一种自动泊车系统
KR102485480B1 (ko) * 2016-05-02 2023-01-05 현대모비스 주식회사 가상 주차선 생성에 의한 주차 지원 장치 및 그 방법
CN106080774A (zh) * 2016-07-13 2016-11-09 重庆长安汽车股份有限公司 引导式泊车系统及方法
CN106541944B (zh) * 2016-11-07 2019-03-05 纵目科技（上海）股份有限公司 一种库位检测方法、系统、及移动设备
CN208069533U (zh) * 2017-12-15 2018-11-09 蔚来汽车有限公司 自动泊车的设备、智能汽车以及计算机存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102933442A (zh) * 2010-06-09 2013-02-13 法雷奥开关和传感器有限责任公司 用于当在停车位中停放时辅助机动车辆的驾驶员的方法、驾驶员辅助装置和机动车辆
CN202923500U (zh) * 2012-11-27 2013-05-08 德尔福电子(苏州)有限公司 电动车辆无线充电引导系统
CN106004515A (zh) 2016-05-12 2016-10-12 广州橙行智动汽车科技有限公司 用于电动汽车自动泊车的车速控制方法及系统
DE102016013180A1 (de) * 2016-11-04 2017-05-24 Daimler Ag Verfahren zum Positionieren eines Kraftfahrzeugs relativ zu einer induktiven Ladeeinrichtung
CN108569280A (zh) * 2017-12-15 2018-09-25 蔚来汽车有限公司 自动泊车的方法和设备、智能汽车以及计算机存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3725626A4

Also Published As

Publication number Publication date
CN108569280B (zh) 2020-03-24
EP3725626B1 (en) 2024-02-07
EP3725626A4 (en) 2021-08-18
CN108569280A (zh) 2018-09-25
EP3725626A1 (en) 2020-10-21

Similar Documents

Publication Publication Date Title
CN109435942B (zh) 一种基于信息融合的车位线车位识别方法及装置
CN102435174B (zh) 基于混合式双目视觉的障碍物检测方法及装置
CN102541063B (zh) 缩微智能车辆寻线控制方法和装置
CN105160884A (zh) 一种融合视频监控和rfid的车辆识别方法及装置
CN103359020B (zh) 一种机动车驾驶培训或考试监控方法和系统
WO2019114731A1 (zh) 自动泊车的方法和设备、智能汽车以及计算机存储介质
CN105488453A (zh) 一种基于图像处理的驾驶员未系安全带检测识别方法
CN103310231A (zh) 车标定位与识别方法
CN102902957A (zh) 一种基于视频流的自动车牌识别方法
CN104916162A (zh) 泊车位检测方法和系统
CN111681283B (zh) 一种基于单目立体视觉的应用于无线充电对位的相对位姿解算方法
CN114347847B (zh) 用于协助泊车的方法、控制装置、存储介质和换电站
CN112885108B (zh) 一种基于深度学习算法的停车位上车辆换车检测方法及系统
CN104299443A (zh) 基于低频有源射频识别技术的停车场车辆定位系统和方法
CN208069533U (zh) 自动泊车的设备、智能汽车以及计算机存储介质
CN115272655A (zh) 用于换电机器人的多类型电池包视觉定位方法与系统装置
CN105913034A (zh) 车辆识别方法、装置及车辆
WO2019114761A1 (zh) 自动泊车的方法和设备、智能汽车以及计算机存储介质
CN105516661A (zh) 鱼眼相机与ptz相机相结合的主从目标监控系统及方法
CN201345166Y (zh) 视频通道车辆检测器
CN104408431A (zh) 交通监控下的车款识别方法
CN103488801A (zh) 一种基于地理信息空间库的机场目标检测方法
CN205905807U (zh) 一种车辆无线充电对位系统
CN106004491A (zh) 一种车辆无线充电对位方法和系统
Zhang et al. Detection algorithm of takeover behavior of automatic vehicles’ drivers based on deep learning

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18888380

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018888380

Country of ref document: EP

Effective date: 20200715