TWI709944B

TWI709944B - Parking management system

Info

Publication number: TWI709944B
Application number: TW108120479A
Authority: TW
Inventors: 黃榮堂
Original assignee: 奇異平台股份有限公司
Priority date: 2019-06-13
Filing date: 2019-06-13
Publication date: 2020-11-11
Also published as: TW202046253A; CN212809464U

Abstract

A parking management system includes a parking space disposed along a lane and a parking space sensor disposed in the parking space, and at least two adjacent license plate detectors are disposed at two ends of the lane, that is, the first license plate detector and the second license plate detector to identify the vehicle license plate entering the lane by the first license plate number detector, and the second license plate number detector identifies the vehicle license plate leaving the lane, the parking space sensing. The device can accurately detect the occupancy of the vehicle in real time, and upload the message of the start and end time of the occupancy to the server or the cloud, and comprehensively judge the vehicle license plate and the vehicle type that are parked in the parking space.

Description

停車管理系統 Parking management system

本發明係關於一種停車管理系統；特別關於一種街道上與非街道停車場的管理系統，該系統係利用安裝於岔路口的車牌號偵測器偵測通過的車輛車牌號並與車位感測器連動，得知毎一停車位停入的車輛車牌號。 The present invention relates to a parking management system; particularly to a management system for on-street and off-street parking lots. The system uses a license plate number detector installed at a fork to detect the license plate number of a passing vehicle and is linked with a parking sensor , Know the license plate number of the vehicle parked in each parking space.

習知的街道上或街道外的停車場，若需要知道在停車格上的車輛其車牌號碼與車種或顏色，通常就是使用固定於車位附近的攝影機，目前在街道上的應用場景，主要是兩個攝影機外加一超音波車位偵測器對一個車位的設置。而街道外，特別是室內停車場，因為有天花板可架設，因此可以是一個攝影機對應兩到三個車位，若是車位安排是車門相鄰，而非頭尾相鄰，則攝影機可有效安排，達到一對二或一對三，但通常因為空間限制多半為一對二；若是停車格為車輛頭尾相接，則攝影機的安排，最多只能一對一。另外對於非室內街道外的停車場，例如機場的戶外停車場，或是園區的戶外停車場，基本上只有進出閘門的車牌辨識，無法得知車輛停在何處，也不方便毎個車位安裝一個攝影機。由此可知，無論街道上或街道外要大量使用攝影機才能有效偵測停在停車格上的車輛車牌，並且也需要空間與建築物的結構才容許如此安排，帶來了成本高，需要大量配電，安裝不易的難題。本發明提出一個改善方案。 In the conventional parking lot on the street or off the street, if you need to know the license plate number and the type or color of the vehicle on the parking grid, you usually use a camera fixed near the parking space. The current application scenarios on the street are mainly two The camera plus an ultrasonic parking detector to set a parking space. Outside the streets, especially indoor parking lots, because there are ceilings that can be erected, one camera can correspond to two to three parking spaces. If the parking spaces are arranged with doors adjacent to each other instead of adjacent to the front and rear, the cameras can be effectively arranged to achieve one. Two-to-two or one-to-three, but usually because of space constraints, it is usually one-to-two; if the parking space is connected to the end of the vehicle, the camera arrangement can only be one-to-one at most. In addition, for non-indoor parking lots, such as outdoor parking lots at airports, or outdoor parking lots in parks, basically only the license plate recognition of the entrance and exit gates is available. It is impossible to know where the vehicle is parked, and it is not convenient to install a camera per parking space. It can be seen that a large number of cameras are used on or off the street to effectively detect the license plates of vehicles parked on the parking grid, and it also requires space and the structure of the building to allow such an arrangement, which brings high costs and requires a lot of Power distribution, difficult to install. The present invention proposes an improvement scheme.

另外，在高速公路收費站，若要達到自動收費，通常都會選用ETC，ETC基本上就是RFID的技術，在車輛上貼上被動式或安裝RFID標籤，在收費站或是在路上架起龍門式RFID讀取器，需要市電供應電源，由於RFID標籤需要讀取器對其充電，整體而言RFID讀取器體積大，價格高，而且並無法精準定位，例如室內停車格上精準定位每一車輛。此外RFID讀取器在路邊停車格在安裝，經濟效益不高，也有定位的問題。因此雖然多數車輛有RFID標籤，然則有效偵測停在停車格上的車輛車牌，仍然不易，或不符經濟效益，本發明也為此提出一個改善方案。 In addition, in expressway toll stations, if you want to achieve automatic toll collection, ETC is usually used. ETC is basically RFID technology. Passive or install RFID tags on vehicles, and set up gantry RFID at toll stations or on the road. The reader needs a mains power supply. Since the RFID tag needs to be charged by the reader, the RFID reader is bulky, expensive, and cannot be accurately positioned, such as accurately positioning each vehicle on an indoor parking grid. In addition, RFID readers are installed in the parking grid on the side of the road, which is not economical and has positioning problems. Therefore, although most vehicles have RFID tags, it is still difficult to effectively detect the license plate of the vehicle parked on the parking grid, or it is not economical. The present invention also proposes an improved solution for this.

本發明的目的在於提出一種停車管理系統，包含設置於一段車道旁的車位與設置於該車位的車位感測器，至少兩台相鄰的車牌號偵測器設置於該段車道的兩端，亦即第一車牌號偵測器與第二車牌號偵測器，藉由第一車牌號偵測器辨識進入該段車道的車輛車牌，第二車牌號偵測器辨識離開該段車道的車輛車牌，該車位感測器可即時精準偵測車輛佔位與否，並上傳佔位起迄時間的訊息給伺服器或雲端，可綜合研判停入車位的車輛車牌與車種。 The object of the present invention is to provide a parking management system, which includes a parking space set beside a section of lane and a parking space sensor set in the parking space. At least two adjacent license plate number detectors are set at both ends of the section of lane. That is, the first license plate number detector and the second license plate number detector, the first license plate number detector recognizes the license plate of the vehicle entering the section of the lane, and the second license plate number detector recognizes the vehicle leaving the section of the lane License plate, the parking sensor can accurately detect whether a vehicle occupies a seat or not, and upload information about the start and end time of the occupancy to the server or the cloud, which can comprehensively determine the license plate and type of the vehicle parked in the parking space.

本發明的目的在於提出一種停車管理系統，包含設置於一段車道旁的車位與設置於該車位的車位感測器，至少兩台相鄰的車牌號偵測器設置於該段車道的兩端，該車牌號偵測器選自一可讀取車牌號的攝影機、可讀取車輛ID標籤(Tag)內車牌號的RFID讀取器、可讀取藍芽電子車牌內車牌號的藍芽讀取器、可讀取WIFI電子車牌內車牌號的WIFI讀取器、可讀取4G/5G電子車牌內車牌號的4G/5G讀取器。 The object of the present invention is to provide a parking management system, which includes a parking space set beside a section of lane and a parking space sensor set in the parking space. At least two adjacent license plate number detectors are set at both ends of the section of lane. The license plate number detector is selected from a camera that can read the license plate number, an RFID reader that can read the license plate number in the vehicle ID tag (Tag), and a Bluetooth reader that can read the license plate number in the Bluetooth electronic license plate. WIFI reader, WIFI reader that can read the license plate number in WIFI electronic license plate, A 4G/5G reader that reads the license plate number in the 4G/5G electronic license plate.

11‧‧‧超音波感測器 11‧‧‧Ultrasonic sensor

12‧‧‧紅外光感測器 12‧‧‧Infrared light sensor

13‧‧‧雷達/微波感測器 13‧‧‧Radar/Microwave Sensor

14‧‧‧紅燈或綠燈 14‧‧‧Red light or green light

16‧‧‧區域控制器 16‧‧‧Area Controller

17‧‧‧中央控制器 17‧‧‧Central Controller

18‧‧‧伺服器或雲端 18‧‧‧Server or Cloud

21‧‧‧安裝於地面上的地磁感測器 21‧‧‧A geomagnetic sensor installed on the ground

23‧‧‧車牌號偵測器 23‧‧‧License plate number detector

25‧‧‧超音波感測器/紅外光感測器/雷達/微波感測器之一加上燈號 25‧‧‧One of the ultrasonic sensor/infrared light sensor/radar/microwave sensor plus light signal

27‧‧‧超音波感測器/紅外光感測器/雷達/微波感測器之一 27‧‧‧One of Ultrasonic Sensor/Infrared Light Sensor/Radar/Microwave Sensor

S41-S47‧‧‧本發明之車牌號與停車格對位方法實施例的七個步驟 S41-S47‧‧‧The seven steps of the embodiment of the method for aligning the license plate number and parking grid of the present invention

S51-S57‧‧‧本發明之車牌號與停車格對位方法另一實施例的七個步驟 S51-S57‧‧‧The seven steps of another embodiment of the method for aligning the license plate number and the parking grid of the present invention

60‧‧‧車輛 60‧‧‧vehicle

61‧‧‧第一車牌號偵測器 61‧‧‧The first license plate number detector

62‧‧‧第二車牌號偵測器 62‧‧‧Second license plate number detector

63‧‧‧右轉第一車牌號偵測器 63‧‧‧Turn right first license plate number detector

64‧‧‧直行第一車牌號偵測器 64‧‧‧The first car license plate number detector

65‧‧‧車道 65‧‧‧lane

66‧‧‧左轉第一車牌號偵測器 66‧‧‧Left turn first license plate number detector

70‧‧‧車位感測器 70‧‧‧Parking Sensor

72‧‧‧螺釘 72‧‧‧Screw

74‧‧‧外殼 74‧‧‧Shell

76‧‧‧地面 76‧‧‧Ground

第1圖係習知的室內停車場管理系統。 Figure 1 is a conventional indoor parking lot management system.

第2圖係本發明的室內停車場管理系統。 Figure 2 shows the indoor parking lot management system of the present invention.

第3圖係本發明的室內停車場平面圖。 Figure 3 is a plan view of the indoor parking lot of the present invention.

第4圖係本發明之車牌號與停車格對位的方法與其演算法的一實施例。 Fig. 4 is an embodiment of the method and algorithm for aligning the license plate number and the parking space of the present invention.

第5圖係本發明之車牌號與停車格對位的方法與其演算法的另一實施例。 Figure 5 is another embodiment of the method and algorithm for the registration of the license plate number and the parking space of the present invention.

第6圖係本發明之結合路口攝影機的街道(on-street)停車系統。 Figure 6 shows the on-street parking system combined with intersection cameras of the present invention.

第7圖係本發明的街道外地磁感測器。 Figure 7 is the off-street magnetic sensor of the present invention.

街道外的停車場通常可分成兩種型態，室內停車場與室外停車場 Off-street parking lots can usually be divided into two types, indoor parking lots and outdoor parking lots

街道外的室內停車位基本上大都為至少一個車道，車道兩旁為停車格，通常大型停車場有多個車道，並組成類似棋盤式的車道，與市區路邊停車基本上相似。 Indoor parking spaces outside the street are basically at least one lane, with parking grids on both sides of the lane. Usually, large parking lots have multiple lanes and form a checkerboard-like lane, which is basically similar to urban roadside parking.

如第1圖所示，習知的室內停車場，主要都是使用超音波測距感測器11，相較於其他感測器而言有其優勢，例如紅外光測距感測器12容易受環境光的干擾，而微波或雷達波測距感測器13功耗較大，成本高。通常這些測距感測器也搭配紅燈與綠燈14來顯示其偵測停車位上有無車輛，讓進入停車場的車主可以看到何處有空位，此外這些感測器的偵測空位的數量可以毎一排停車位為群組，搭配LED空位數量顯示器與區域控制器16，最後再匯集到中央控制器17，然後上傳伺服器或雲端18。另外值得注意的地方，乃是上述的測距感測器用於偵測車輛在席與否，也可進一步改由安裝攝影機來取代，同時偵測在席與車牌號。 As shown in Figure 1, conventional indoor parking lots mainly use ultrasonic ranging sensors 11, which have advantages over other sensors. For example, infrared light ranging sensors 12 are easily The interference of ambient light, and the microwave or radar wave ranging sensor 13 has high power consumption and high cost. Usually these range-finder sensors are also combined with red and green lights 14 to show whether there is a car in the parking space, so that car owners entering the parking lot can see where there are empty spaces. In addition, the number of empty spaces detected by these sensors can be Each row of parking spaces is grouped, with LED vacancy quantity indicator and area controller 16, and finally collected to the central controller 17, and then uploaded to the server or cloud 18. Another point worth noting is that the above-mentioned range sensor is used to detect the presence or absence of a vehicle. It can also be replaced by a camera to detect the presence and license plate number.

如第2圖所示，為本發明的室內停車場管理系統，主要是採用安裝於地面上的地磁感測器21，來取代第1圖所示各種安裝於天花板上的測距感測器，這些地磁感測器21優選的實施方式是使用電池，內建低功耗藍芽通訊，若必要也可再加入WIFI，主要用於將車位在席與否的資訊上傳至附近的車牌號偵測器23，該車牌偵測器23內建藍芽與WIFI，扮演路由器的功能，而WIFI也可由乙太網路或LPWA或3G/4G/5G來替代，用於聯接伺服器或雲端，因此車位在席與否，以及路經車牌號偵測器23底下的車輛車牌號等資訊都可上傳至伺服器或雲端(未顯示於圖中)。在街道外的停車場，不會受到路過大型卡車與客車的干擾，因此地磁感測器21可以僅藉由車輛停放於地磁感測器上方對於磁場偵測的變化，就可精準判斷車輛的在席與否，成本低廉，加上地磁感測器的功耗很低，又使用藍芽低功耗通訊，僅在車停入駛離，才會啟動通訊，所需時間通常可以數秒以內，估計每天每一停車格每天最多停入駛離的次數不會超過20次，因此一個6Ah的電池，可望至少5年不必更換。 As shown in Figure 2, the indoor parking lot management system of the present invention mainly uses the geomagnetic sensor 21 installed on the ground instead of the various distance measuring sensors installed on the ceiling shown in Figure 1. These The preferred implementation of the geomagnetic sensor 21 is to use a battery, built-in low-power Bluetooth communication, if necessary, can also add WIFI, mainly used to upload the information of whether the car is seated or not to the nearby license plate number detector 23. The license plate detector 23 has built-in Bluetooth and WIFI, acting as a router, and WIFI can also be replaced by Ethernet or LPWA or 3G/4G/5G to connect to the server or the cloud, so the parking space is Information such as whether it is seated or not, and the license plate number of the vehicle passing under the license plate number detector 23 can be uploaded to the server or the cloud (not shown in the figure). The parking lot outside the street will not be disturbed by large trucks and buses passing by. Therefore, the geomagnetic sensor 21 can accurately determine the presence of the vehicle only by detecting changes in the magnetic field just by parking the vehicle above the geomagnetic sensor. It is low cost or not, coupled with the low power consumption of the geomagnetic sensor, and Bluetooth low-power communication, only when the car is parked or left, the communication will be started. The time required can usually be within a few seconds, estimated every day The maximum number of parking in and out of each parking space per day will not exceed 20 times, so a 6Ah battery is expected to not need to be replaced for at least 5 years.

如第2圖所示，在另一實施例，安裝於地面上的地磁感測器21，也可沿用超音波感測器、紅外光感測器、雷達、微波感測器之一加上燈號25，或是沒有燈號，只有超音波感測器、紅外光感測器、雷達、微波感測器之一27。 As shown in Figure 2, in another embodiment, the geomagnetic sensor 21 installed on the ground can also use one of ultrasonic sensors, infrared light sensors, radar, and microwave sensors. Light 25, or no light, only one of ultrasonic sensor, infrared light sensor, radar, microwave sensor 27.

如第3圖所示，為一典型室內停車場的平面圖，有些車道容許單輛車行走，有些容許雙向車行走，實際上，車主有可能不遵行方向來走，因此對於這類車輛，要能判讀前車牌或是後車牌。如圖2所示的車牌號偵測器23，可安裝於在車道上方的天花板，或固定或垂吊，該車牌號偵測器23，可以有效偵測通過其底下車輛的車牌號。該車牌號偵測器23的實施方式可以有多種，選自一可讀取車牌號的攝影機、可讀取車輛ID標籤(Tag)內車牌號的RFID讀取器、可讀取藍芽電子車牌內車牌號的藍芽讀取器、可讀取WIFI電子車牌內車牌號的WIFI讀取器、可讀取4G/5G電子車牌內車牌號的4G/5G讀取器。值得提起的是上述的電子車牌，可以是車主的行動裝置，例如手機。 As shown in Figure 3, it is a floor plan of a typical indoor parking lot. Some lanes allow a single vehicle to travel, and some allow two-way vehicles. In fact, the owner may not follow the direction. Therefore, for this type of vehicle, it must be able to interpret Front license plate or rear license plate. The license plate number detector 23 shown in FIG. 2 can be installed on the ceiling above the driveway, or fixed or suspended. The license plate number detector 23 can effectively detect the license plate numbers of vehicles passing underneath it. The license plate number detector 23 can be implemented in multiple ways, selected from a camera that can read the license plate number, an RFID reader that can read the license plate number in a vehicle ID tag (Tag), and a Bluetooth electronic license plate that can read Bluetooth reader for internal license plate number, WIFI reader that can read license plate number in WIFI electronic license plate, 4G/5G reader that can read license plate number in 4G/5G electronic license plate. It is worth mentioning that the above-mentioned electronic license plate can be a mobile device of the car owner, such as a mobile phone.

如圖3所示，安裝這些車牌號偵測器23的地點，原則就是在進出口，以及岔路口，以及每間隔16-20個車位，如此可確保進入兩個相鄰的車牌號偵測器23之間的車輛，無論停入其間的停車位，或離開其間的停車位，其車牌號都會受到該兩個相鄰的車牌號偵測器23至少其中之一的辨識與偵測。以下以攝影機來實施車牌號偵測器23，詳加說明其構造與方法。 As shown in Figure 3, the location where these license plate number detectors 23 are installed should be at the entrances and exits, and forks, and at intervals of 16-20 parking spaces, so as to ensure access to two adjacent license plate number detectors. For vehicles between 23, no matter they park in the parking space in between, or leave the parking space in between, their license plate numbers will be recognized and detected by at least one of the two adjacent license plate number detectors 23. In the following, a camera is used to implement the license plate number detector 23, and its structure and method are described in detail.

街道外的停車場特別室內，在影像處理方面相較於街道上的路邊停車要單純一點，不會有霧霾，光線較充足與穩定，車輛的車速也較慢許多，這些地方的車道通常最多只容許一來一往雙向，單一方向同一時間只會有一輛通過攝影機，加上車速約略落在20公里/HR，因此每秒約行進5公尺，若攝影機視野範圍為5公尺，則即使每秒取樣兩個畫面，應該至少會拍到有明顯車牌的影像，如此只要利用車牌的搜尋比對，即可取得具有車牌的影像，並將此以4G或是WIFI 6.0快速上傳送至雲端進行辨識。整體處理時間應可以在0.1-2秒內完成，一般車距若是落在5米以上，應可以有效達成辨認車牌的需求。進一步，若使用第五代通訊5G，上傳影像至雲端，所需時間更可縮減至十毫秒以內，雲端進行AI的深度學習判識，整體時間最多在數十毫秒內完成；又或是直接在攝影機的運算器內完成AI車牌辨識，只需要運算器的運算能力提升即可。 The parking lot outside the street is especially indoor. Compared with the roadside parking on the street, the image processing is simpler, there is no haze, the light is sufficient and stable, and the speed of the vehicle is much slower. The lanes in these places are usually the most. Only one direction and one direction are allowed. Only one camera will pass through the camera at the same time in a single direction. In addition, the vehicle speed is about 20 kilometers/HR, so it travels about 5 meters per second. If the camera's field of view is 5 meters, even Sampling two frames per second, it should be It is rare to capture images with obvious license plates, so as long as the search and comparison of license plates are used, images with license plates can be obtained, and the images can be quickly sent to the cloud with 4G or WIFI 6.0 for identification. The overall processing time should be completed within 0.1-2 seconds. Generally, if the distance between vehicles is more than 5 meters, it should be able to effectively meet the needs of identifying license plates. Furthermore, if the fifth-generation communication 5G is used to upload images to the cloud, the time required can be reduced to less than ten milliseconds. The cloud performs AI deep learning and identification, and the overall time can be completed within tens of milliseconds at most; or directly The AI license plate recognition is completed in the arithmetic unit of the camera, and only the computing power of the arithmetic unit is improved.

另一種實施例，一般車牌辨識，需要能拍攝整個車身，因此，就能辨識車尾或車頭的車牌，這只要是利用高速的運算器，可以依照毎秒30個畫面，去進行毎個畫面的運算，搭配深度學習的演算法，即可快速辨識車牌。例如，2018年英特爾公司的研究人員提出的：LPRNet：License Plate Recognition via Deep Neural Networks(http：//cn.arxiv.org/pdf/1806.10447)，判斷一個車牌號，搭配不同的運算器，例如Rasbberry Pi***USB的OpenVino AI加速器，所需時間落在幾毫秒以內，體積小，含攝影機整個成本落在200-300USD。其他演算法也可行，例如github上的開源方案包括OpenALPR，easyPR，HyperLPR等。 In another embodiment, general license plate recognition needs to be able to photograph the entire vehicle body. Therefore, the license plate at the rear or front of the vehicle can be recognized. As long as the high-speed calculator is used, the calculation of each frame can be performed according to 30 frames per second. , With the deep learning algorithm, you can quickly identify the license plate. For example, in 2018, Intel’s researchers proposed: LPRNet: License Plate Recognition via Deep Neural Networks (http://cn.arxiv.org/pdf/1806.10447) to determine a license plate number and match it with different computing devices, such as Rasbberry Pi plugs into the USB OpenVino AI accelerator, the time required falls within a few milliseconds, the size is small, and the entire cost including the camera is 200-300USD. Other algorithms are also feasible. For example, open source solutions on github include OpenALPR, easyPR, HyperLPR, etc.

雖然上述的使用情境，在正常狀況下，單一車輛進入第一與第二車牌號偵測器之間，都可有效將停入停車位的車輛其車牌號正確無誤的指派，然則在某些實施例，可能第一與第二車牌號偵測器之間，會有多輛車先後進入這個區域，但可能車主有所猶豫，先進者並未先停或是停車技術不熟練，沒有一次到位，無法完全靠先進者先停好車的常規。因此就需要另外利用車主APP來確認，該車主APP可以讓：當停入停車格後，車主的手機藍芽可與地面的車位感測器藍芽連線，取得車位編碼或是將該車位所在位置的經緯高度上傳雲端，或是連上網路報到，讓雲端取得停入該車位的車輛車牌號。另外，也可讓車位感測器藉由與車主手機連線取得車輛車牌號後，利用藍牙上傳給車牌號偵測器，告知停入者的車牌號，車牌號偵測器再上傳雲端，即可100%確定該車牌號的車輛停入的位置。再者，若車主沒有安裝或執行車主APP，雲端若可透過車牌號取得其手機號碼或EMAIL地址，也可由手機發簡訊，或EMAIL告知其是否停入某一車位編號的停車位，或給予該室內停車場地圖的停車位置，並請其確認，也可100%確定該車牌號的車輛停入的位置。 Although the above usage scenarios, under normal conditions, a single vehicle entering between the first and second license plate number detectors can effectively assign the correct license plate number of the vehicle parked in the parking space, but in some implementations For example, it is possible that between the first and second license plate number detectors, there will be multiple vehicles entering this area one after another, but the owner may hesitate. It is impossible to completely rely on the rule of the first person to park the car first. Therefore, it is necessary to additionally use the owner's APP to confirm, the owner's APP can let: when parked in the parking space, The Bluetooth of the main mobile phone can be connected to the Bluetooth of the parking space sensor on the ground to obtain the parking space code or upload the latitude and longitude height of the parking space to the cloud, or connect to the Internet to report to the cloud to get the car parked in the parking space number plate. In addition, you can also allow the parking sensor to obtain the license plate number of the vehicle by connecting with the owner's mobile phone, and upload it to the license plate number detector using Bluetooth to inform the parking person's license plate number. The license plate number detector uploads it to the cloud. It can be 100% determined where the vehicle with the license plate number is parked. Furthermore, if the car owner has not installed or executed the owner’s APP, if the cloud can obtain his mobile phone number or email address through the license plate number, he can also send a text message from his mobile phone, or email him to tell him whether to park in a parking space with a certain parking number, or give the The parking location of the indoor parking lot map, and ask it to confirm, it can also be 100% determined where the vehicle with the license plate number will park.

其實，本發明的車主APP可提供以下優點： In fact, the owner APP of the present invention can provide the following advantages:

1.在室內獲得車牌號偵測器廣播的經緯高度，進行定位，獲得導航抵達指定停車格。 1. Obtain the latitude and longitude height broadcast by the license plate number detector indoors, perform positioning, and obtain navigation to reach the designated parking space.

2.車主只要按下APP上[自主報到]的按鍵，直接扣款，可享以分計費，打折付帳，無需去超商繳費，也無需下車在POS機繳費。 2. As long as the car owner presses the button of [self-report] on the APP, the payment can be directly deducted, and the bill will be paid in points, discounted, and there is no need to go to the supermarket to pay, and there is no need to get off the car to pay at the POS machine.

3.可容易反向尋車，只要打開車主APP，一進入停車場，立即取得車牌號偵測器發送的經緯高度，手機即可定位自身，同時顯示停車場的地圖，以及愛車的位置，因此車主可接受導航，前往愛車的位置。 3. It is easy to find the car in the reverse direction. As long as you open the owner's APP, once you enter the parking lot, you can immediately obtain the latitude and longitude height sent by the license plate number detector. Accept the navigation and go to the location of your car.

4.可預約停車位，並透過導航，即可達成定時定點抵達預約的停車位。 4. Parking space can be reserved, and through navigation, you can reach the reserved parking space at a fixed time.

除了上述室內停車管理系統外，結合路口攝影機的非街道(off-street)停車與與其停車系統，如圖3所示，非街道的停車系統，基本上會有至少一個車道，車道兩旁或至少一旁是停車位，任何一個車道，都會有出入口，無論為單向或雙向，只需要在該車道的轉向口或岔路口各分別安裝一個車牌號偵測器，此處以攝影機為例，該對攝影機面對面設置，所以可辨識相鄰兩個轉向口或岔路口之間車道出入的車輛及其車牌，等效上，這段車道與其兩旁的停車位，可視同為一個封閉的停車場，僅有兩個出入口，因此在出入口設置攝影機就可以進行習知大型停車場管理閘口出入，使用攝影機辨識車牌的方法，並上傳雲端，然則，本發明，進一步要使用在席車位感測器，來判斷經過入口的車輛，進入車道旁的停車格停入的位置，並給予對應的車牌，如此，無需在每兩或三個相鄰車位安裝一部攝影機，來監控停入的車輛其車牌。停入的車輛，經過一段時間，若是駛離該車位，車位感測器也可偵測該車輛離開，同時路口的攝影機又會再度看到該車輛的車牌。 In addition to the above-mentioned indoor parking management system, the off-street parking and its parking system combined with the intersection camera, as shown in Figure 3, the off-street parking system basically has at least one lane, on both sides or at least one side of the lane It’s a parking space, any lane, there will be For entrances and exits, whether one-way or two-way, you only need to install a license plate number detector at the turn or fork of the lane. Take the camera as an example. The pair of cameras are set face to face, so two adjacent turns can be identified Vehicles entering and exiting the lanes or forks and their license plates, equivalently, this section of the lane and the parking spaces on both sides can be regarded as a closed parking lot with only two entrances and exits. Therefore, you can set up cameras at the entrances and exits for learning. Knowing the method of entering and exiting the management gate of a large parking lot, using the camera to recognize the license plate and uploading it to the cloud, however, the present invention further uses the parking space sensor to determine the position of the vehicle passing the entrance and entering the parking space next to the lane. , And give the corresponding license plate, so there is no need to install a camera in every two or three adjacent parking spaces to monitor the license plate of the parked vehicle. If a parked vehicle leaves the parking space after a period of time, the parking sensor can also detect that the vehicle is leaving, and the camera at the intersection will see the license plate of the vehicle again.

參考第4圖，本發明之車牌號與停車格對位的方法與其演算法的實施例，如下列所述 Referring to Fig. 4, the method for aligning the license plate number with the parking grid and the algorithm of the embodiment of the present invention are as follows

步驟S41. 比較從第一攝影機進入的車牌號與從第二攝影機出去的車牌號。若某一車牌號有進入[選定車道]的紀錄，但經過一個時間差，該車牌號未有離開[選定車道]的紀錄，則將該車牌號的進入時間Tin，加以紀錄。該車牌號離開的時間Tout也加以記錄。若某一車牌號一進一出，而且相差時間在一個範圍內，則該車並未停入其間的車位，可加以忽略。 Step S41. Compare the license plate number entered from the first camera with the license plate number exited from the second camera. If a certain license plate number has a record of entering [selected lane], but after a time difference, the license plate number does not have a record of leaving [selected lane], then the entry time Tin of the license plate number is recorded. The time Tout of the license plate number is also recorded. If a certain license plate number enters and exits, and the time difference is within a range, the car has not parked in the parking space in between and can be ignored.

步驟S42. 停車格上的在席車位感測器若被觸發確認有車停入，紀錄該觸發時間Tsin。 Step S42. If the parking space sensor on the parking grid is triggered to confirm that a car has parked, record the trigger time Tsin.

步驟S43. 雲端比較Tsin與Tin，若0<(Tsin-Tin)<Th則該新停入的車的車牌號，即為步驟S41於Tin取得的車牌號。如果只有Tin，沒有Tsin，則表示該車牌號違規停車，未停入車格內，或是仍未決定是否停入。 Step S43. Compare Tsin and Tin in the cloud. If 0<(Tsin-Tin)<Th, the license plate number of the newly parked car is the license plate number obtained in Tin in step S41. If only Tin, no With Tsin, it means that the license plate number has illegally parked, has not parked in the car compartment, or has not yet decided whether to park.

步驟S44. 若車位已滿，並且如果只有Tin，沒有Tsin，則表示該車牌號違規停車，未停入車格內。 Step S44. If the parking space is full, and if there is only Tin but no Tsin, it means that the license plate number has illegally parked and has not been parked in the car compartment.

步驟S45. 若是確認停車，則該車牌號，會經過一段時間，出現在車牌號離開的名單上，Tout。 Step S45. If parking is confirmed, the license plate number will appear on the list of license plate number departures after a period of time, Tout.

步驟S46. 停車格上的感測器若被觸發確認有車駛離，紀錄該觸發時間Tsout。 Step S46. If the sensor on the parking grid is triggered to confirm that a car is leaving, record the trigger time Tsout.

步驟S47. 雲端比較Tsout與Tout，若0<(Tout-Tsout)<Th則該新離開的車的車牌號，即為步驟S45於Tout取得的車牌號。如果只有Tout沒有Tsout，則表示該車牌號違規停車，未停入車格內。 Step S47. The cloud compares Tsout and Tout. If 0<(Tout-Tsout)<Th, the license plate number of the newly leaving vehicle is the license plate number obtained at Tout in step S45. If there is only Tout and no Tsout, it means that the license plate number is illegally parked and not in the compartment.

參考第5圖，本發明之車牌號與停車格對位的方法與其演算法的另一實施例，步驟S51. 第一攝影機紀錄毎一進入車輛的車牌號num與時間Tin，C(num,Tin)，並上傳至雲端；而第二攝影機紀錄毎一出去車輛的車牌號num與時間Tout，C(num,Tout)，並上傳至雲端。 Referring to Fig. 5, another embodiment of the method for aligning the license plate number and the parking space and its algorithm of the present invention, Step S51. The first camera records the license plate number num and time Tin,C(num,Tin) of each entering vehicle and uploads it to the cloud; while the second camera records the license plate number num and time Tout,C( num,Tout) and upload to the cloud.

步驟S52. 停車格上的在席車位感測器若被觸發確認有車停入，紀錄該觸發時間Tsin，將該地點與Tsin上傳至雲端。 Step S52. If the parking space sensor on the parking grid is triggered to confirm that a car is parked, record the trigger time Tsin, and upload the location and Tsin to the cloud.

步驟S53. 雲端根據步驟S52取得信息的地點，找到對應的第一攝影機與第二攝影機的紀錄，並藉由步驟S52取得的Tsin，比較Tsin與Tin，若0<(Tsin-Tin)<Th則該新停入的車的車牌號，即為步驟1於Tin取得的車牌號。 Step S53. The cloud finds the corresponding records of the first camera and the second camera according to the location where the information is obtained in step S52, and compares Tsin and Tin based on the Tsin obtained in step S52, if 0<(Tsin-Tin)<Th then The license plate number of the newly parked car is the license plate number obtained in Tin in step 1.

步驟S54. 若車位已滿，並且如果只有Tin，沒有Tsin，則表示該車牌違規停車，未停入車格內。 Step S54. If the parking space is full, and if there is only Tin but no Tsin, then the table Shows that the license plate has parked illegally and has not parked in the compartment.

步驟S55. 若是確認停車，則該車牌號，會經過一段時間，出現在車牌號離開的名單上，Tout。 Step S55. If parking is confirmed, the license plate number will appear on the list of license plate number leaving after a period of time, Tout.

步驟S56. 停車格上的感測器若被觸發確認有車駛離，紀錄該觸發時間Tsout。 Step S56. If the sensor on the parking grid is triggered to confirm that a car is leaving, record the trigger time Tsout.

步驟S57. 雲端比較Tsout與Tout，若0<(Tout-Tsout)<Th則該新離開的車的車牌號，即為步驟S55於Tout取得的車牌號。如果只有Tout，沒有Tsout，則表示該車牌號違規停車，未停入車格內。 Step S57. The cloud compares Tsout and Tout. If 0<(Tout-Tsout)<Th, the license plate number of the newly leaving car is the license plate number obtained at Tout in step S55. If there is only Tout and no Tsout, it means that the license plate number has illegally parked and has not been parked in the compartment.

需要注意的是，優選的實施例，是讓車位感測器具備有藍芽或WIFI或LPWA的無線通訊能力，而第一攝影機與第二攝影機則同樣具有藍芽或WIFI或LPWA的通訊能力。如此有助於讓車位感測器低功耗，並且能與車主的手機上的BLE或WIFI互通。 It should be noted that, in the preferred embodiment, the parking sensor has Bluetooth or WIFI or LPWA wireless communication capabilities, while the first camera and the second camera also have Bluetooth or WIFI or LPWA communication capabilities. This helps to make the parking sensor low power consumption, and can communicate with the BLE or WIFI on the owner's mobile phone.

再者，第一攝影機進一步可安裝一BLE作為定位用的信標(BEACON)，不斷廣播其所在地的經緯高度。此舉可讓車主的手機APP能夠進行停車場的導航，並且能夠直接指派特定車位給預約車主，或是VIP等。 Furthermore, the first camera can be further equipped with a BLE as a positioning beacon (BEACON) to continuously broadcast the latitude and longitude of its location. This will allow car owners’ mobile apps to navigate the parking lot and directly assign specific parking spaces to reserved car owners or VIPs.

在席車位感測器選自安裝於地表面之上或是與地面平或之下的地磁感測器、超音波感測器、紅外線感測器、微波感測器，或其組合。也可選用安裝於天花板的超音波感測器、紅外線感測器、微波感測器，或其組合。 The seat parking sensor is selected from a geomagnetic sensor, an ultrasonic sensor, an infrared sensor, a microwave sensor, or a combination thereof, which is installed on the ground surface or is level or below the ground. An ultrasonic sensor, an infrared sensor, a microwave sensor, or a combination thereof installed on the ceiling can also be used.

判識車牌一般需要三道手續，第一、車輛辨識；第二、車牌辨識；第三、OCR文字辨識。在路邊停車，要有效取得車輛的種類，例如機車、卡車、大客車、小客車、小貨車等，非常複雜，經常需要較長的時間來判斷。 Recognition of license plates generally requires three procedures, first, vehicle identification; second, license plate recognition; third, OCR text recognition. To park on the roadside, it is necessary to effectively obtain the types of vehicles, such as locomotives, trucks, buses, small buses, small trucks, etc. It is very complicated and often requires longer Time to judge.

室內停車，通常只限制小客車、小貨車停入，因此有時可以無需第一、車輛辨識的手續。只需要車牌辨識與OCR文字辨識，然則為了增加判識的速度，並減少不必要的取像，可以進一步使用如下所述的車輛通過感測器。 Indoor parking usually only restricts the parking of small passenger cars and small trucks, so sometimes the first, vehicle identification procedures may be unnecessary. Only license plate recognition and OCR text recognition are required. However, in order to increase the recognition speed and reduce unnecessary imaging, the vehicle passing sensor described below can be further used.

本停車系統，其中的第一攝影機與第二攝影機，進一步搭配車輛通過感測器，用來判斷車輛經過攝影機的時間，並觸發攝影機取像。此處的車輛通過感測器可選自安裝於天花板的超音波感測器、紅外線感測器、微波感測器，或其組合，或選自安裝於地表面之上或是與地面平或之下的地磁感測器、超音波感測器、紅外線感測器、微波感測器，或其組合。 In the parking system, the first camera and the second camera are further matched with the vehicle passing sensor to determine the time the vehicle has passed the camera and trigger the camera to take an image. The vehicle passing sensor here can be selected from an ultrasonic sensor, an infrared sensor, a microwave sensor, or a combination thereof installed on the ceiling, or can be selected from being installed on the surface of the ground or flat or flat with the ground. The geomagnetic sensor, ultrasonic sensor, infrared sensor, microwave sensor, or a combination thereof underneath.

在某些優選的實施例中，第一攝影機與第二攝影機，可使用單板電腦，例如樹莓派；必要時可加裝AI加速晶片，例如Intel的Movidius，或是Nvidia的TX2等，加上攝影機，即可執行深度學習的AI影像辨識神經網路，辨識車牌完成前述的三道手續，通常在數毫秒以內，因此對於室內或室外車道上行駛的車輛與車牌辨識，都足以勝任。 In some preferred embodiments, the first camera and the second camera can use a single-board computer, such as a Raspberry Pi; if necessary, an AI acceleration chip can be installed, such as Intel's Movidius, or Nvidia's TX2, etc. With the camera, the AI image recognition neural network of deep learning can be executed to recognize the license plate to complete the aforementioned three procedures, usually within a few milliseconds. Therefore, it is sufficient for the recognition of vehicles and license plates on indoor or outdoor lanes.

結合路口攝影機的街道(on-street)停車與與其停車系統 On-street parking and its parking system combined with intersection cameras

如第6圖所示，車輛60沿著車道65往前直行通過第一車牌號偵測器61，若是沒有出現在第二車牌號偵測器62，同時在第一車牌號偵測器61與第二車牌號偵測器62的路邊停車位的車位感測器被觸發，則可以推定該車位停入的車輛，其車牌號就是不久前出現於第一車牌號偵測器61的車牌號。若是該車輛60出現於第二車牌號偵測器62，並且沒有第一車牌號偵測器61與第二車牌號偵測器62的路邊停車位的車位感測器被觸發，則該車輛60並未停入路邊停車格。該車輛有可能持續進入直行的第一車牌號偵測器64或進入右轉第一車牌號偵測器63或進入左轉第一車牌號偵測器66。由這個實施情境，無論街道路線多複雜，對於路邊停車位的管理，就是在設有路邊停車位的起始路口安裝第一車牌號偵測器，而在路邊停車位的終止岔路口則要安裝第二車牌號偵測器，搭配車位感測器，就可有效將停入停車格的車輛其車牌號精確指派出來，無需在每個停車格都設立一個車牌號偵測器，例如攝影機，大幅增加設置成本以及之後的維修成本。需要注意的是只要車道沒有路邊停車，就不需要安裝車牌號偵測器。 As shown in Figure 6, the vehicle 60 goes straight ahead along the lane 65 and passes the first license plate number detector 61. If it does not appear on the second license plate number detector 62, at the same time the first license plate number detector 61 and If the parking sensor of the on-street parking space of the second license plate number detector 62 is triggered, it can be inferred that the vehicle parked in the parking space has the license plate number that appeared in the first license plate number detector 61 not long ago. . If the vehicle 60 appears in the second license plate number detector 62, and the parking space sensor of the on-street parking space without the first license plate number detector 61 and the second license plate number detector 62 is triggered, the The vehicle 60 has not parked in the roadside parking space. The vehicle may continue to enter the first license plate number detector 64 that is going straight or enter the first right-turn first license plate number detector 63 or enter the first-turn left license plate number detector 66. From this implementation scenario, no matter how complex the street route is, the management of on-street parking spaces is to install the first license plate number detector at the starting intersection with on-street parking spaces, and at the ending fork of the roadside parking spaces It is necessary to install a second license plate number detector, together with the parking space sensor, to effectively assign the license plate number of the vehicles parked in the parking space, without setting up a license plate number detector in each parking space, for example The camera greatly increases the cost of setup and subsequent maintenance. It should be noted that as long as there is no roadside parking in the driveway, there is no need to install a license plate number detector.

綜上所述，街道上任何一個路段，只要開放路邊停車，並且其出入口只有一個，則只需要在毎個停車位安裝車握感測器，同時在該路口各分別安裝一個攝影機，辨識出入的車輛其車牌，並上傳雲端，即可將毎一停入其間停車位的車輛對應車牌號加以對位，其詳細的演算法流程如第4圖與第5圖所述，在此不再贅述。 In summary, on any section of the street, as long as the roadside parking is open and there is only one entrance and exit, you only need to install a car grip sensor in each parking space, and install a camera at each intersection to identify the entrance and exit. The license plate of each vehicle parked in the parking space can be aligned by uploading the license plate to the cloud. The detailed algorithm flow is as described in Figure 4 and Figure 5, and will not be repeated here. .

相較於室內停車場而言，車牌號偵測器使用攝影機來實施，需要注意的是在街道上架設攝影機，在天候良好，能見度高的條件下，可以取得很好的車牌號辨識效果，然則由於光照、大霧、沙塵暴等天氣影響，導致採集到的圖像品質較低，字元受到雜訊干擾，造成部分目標圖像被背景圖像掩蓋，降低車牌字元識別的準確率。因此本發明也提出另一種對策，在某些國家例如台灣，由於大多數車輛都有貼上eTag，此即是RFID被動式標籤，因此可使用這個ETC的技術搭配攝影機，在路口安裝RFID讀取器於攝影機或單獨使用，等同電子收費站的架構，只要帶有RFID被動式標籤的車輛經過，都可被偵測，偵測率為100%。因此利用本發明的車牌號偵測器不限於攝影機，也涵蓋ETC的技術，甚至有些國家或地區有電子車牌，則車牌號偵測器可以是可讀取藍芽電子車牌內車牌號的藍芽讀取器、或可讀取WIFI電子車牌內車牌號的WIFI讀取器、或可讀取4G/5G電子車牌內車牌號的4G/5G讀取器。 Compared with the indoor parking lot, the license plate number detector is implemented with a camera. It should be noted that setting up a camera on the street can achieve a good license plate number recognition effect under good weather and high visibility conditions. The influence of weather such as light, heavy fog, sandstorm and other weather results in low quality of the collected images, and the characters are interfered by noise, causing part of the target image to be covered by the background image, reducing the accuracy of license plate character recognition. Therefore, the present invention also proposes another countermeasure. In some countries such as Taiwan, since most vehicles have eTags affixed, this is an RFID passive tag. Therefore, this ETC technology can be used with cameras to install RFID readers at intersections. Used in a camera or alone, it is equivalent to the structure of an electronic toll station. As long as a vehicle with an RFID passive tag passes by, it can be detected with a detection rate of 100%. Therefore, the license plate number of the present invention The detector is not limited to cameras, but also covers ETC technology. Even some countries or regions have electronic license plates, the license plate number detector can be a Bluetooth reader that can read the license plate number in the Bluetooth electronic license plate, or readable Take a WIFI reader for the license plate number in the WIFI electronic license plate, or a 4G/5G reader that can read the license plate number in the 4G/5G electronic license plate.

實施例一室內停車場的地磁感測器 Example 1 Geomagnetic sensor for indoor parking lot

如第7圖所示，室內停車場的車位感測器70優選地磁感測器(未示出)來偵測車輛的有無停入，對於車輛而言，因為地磁感測器相當低廉，並且在室內停車格，地磁不易受到外界干擾，特別是室內(街道外)停車場通常限停小客車，這些小客車經過街道上與街道外的實驗，證明當其在車道行駛通過車位上的地磁感測器，由於距離超過3米左右，這樣的旁車對地磁的干擾，遠比街道上有大客車與大卡車經過車位上的地磁感測器，要小的很多，可以說幾乎不會影響地感測器的感測值。此外地磁感測器的功耗很低，可以安裝於地面上，如第7圖所示，使用3-4個螺釘72，即可固鎖於地面76，無需於地面上挖洞，不會對地面造成破壞。地磁感測器對於非金屬的外殼74都可有效穿透，其高度可在20mm以下，直徑可在200mm以下。內建一微控制器(MCU)，即可正確取得地磁的動態數值，並經由適當挑選電池，即可五年內無需更換電池。綜合而言，安裝於室內(街道外)停車場的地磁感測器，其組成包含一3D磁場感測器，一個含有藍芽通訊功能的微控制器(MCU)例如TI CC2642或是TI CC2640等，一電池(可使用鹼性電池，成本更低廉)，一電路板以承載上述元件，一外殼，因此成本相當低，相較於習知安裝於天花板的超音波、紅外光、微波感測器，而言，不但成本有競爭力，同時體積小，功耗低。特別是由於許多室內停車場的天花板高度不一，為了架設超音波、紅外光、微波感測器等，需要佈設懸吊架，以及配置電源線等，成本高，維護費較高。 As shown in Figure 7, the parking space sensor 70 of the indoor parking lot is preferably a magnetic sensor (not shown) to detect the presence or absence of a vehicle. For vehicles, because the geomagnetic sensor is quite cheap and indoor In parking spaces, geomagnetism is not susceptible to external interference. Especially indoor (off-street) parking lots are usually restricted to small passenger cars. These small passenger cars have been tested on and off the street and proved that when they drive in the lane and pass the geomagnetic sensor on the parking space, Because the distance is more than 3 meters, the interference of such sidecars on the geomagnetic field is much smaller than that of the geomagnetic sensor on the parking space with buses and trucks passing by on the street. It can be said that it will hardly affect the geomagnetic sensor. The sensed value. In addition, the geomagnetic sensor has very low power consumption and can be installed on the ground. As shown in Figure 7, using 3-4 screws 72, it can be fixed to the ground 76. There is no need to dig holes in the ground and will not The ground caused damage. The geomagnetic sensor can effectively penetrate the non-metal housing 74, and its height can be less than 20mm, and its diameter can be less than 200mm. A built-in microcontroller (MCU) can correctly obtain the dynamic value of the geomagnetism, and by properly selecting the battery, there is no need to replace the battery within five years. In summary, the geomagnetic sensor installed in the indoor (off-street) parking lot consists of a 3D magnetic field sensor, a microcontroller (MCU) with Bluetooth communication function, such as TI CC2642 or TI CC2640, etc. A battery (alkaline batteries can be used, which is cheaper), a circuit board to carry the above components, and a housing, so the cost is quite low. Compared with conventional ultrasonic, infrared, and microwave sensors installed on the ceiling, In terms of cost, it is not only cost competitive, but also small in size and low power consumption. Especially since the ceiling height of many indoor parking lots is not First, in order to set up ultrasonic, infrared light, microwave sensors, etc., it is necessary to lay a suspension frame and configure a power cord, etc., which is costly and high maintenance.

實施例二街道外露天停車場的地磁感測器 Example 2 Geomagnetic sensor for outdoor parking lot outside the street

街道外露天停車場的地磁感測器基本上與室內停車場的地磁感測器的組成相似，差別在於可以增添太陽能電池，改用充電電池，可以讓更換電池的時間更超過5年。 The composition of the geomagnetic sensor of the outdoor parking lot outside the street is basically similar to that of the indoor parking lot. The difference is that solar cells can be added and rechargeable batteries can be used, which can make the battery replacement time more than 5 years.

實施例三街道上路邊停車位的地磁感測器 Example 3 Geomagnetic sensor for on-street parking spaces on the street

街道上路邊停車位的地磁感測器基本上與街道外露天停車場的地磁感測器的組成相似，差別在於可能增加紅外線測距感測器，以補強地磁感測器容易受旁車經過的干擾。 The composition of the geomagnetic sensor of the on-street parking spaces on the street is basically similar to that of the open-air parking lot off the street, except that an infrared range sensor may be added to strengthen the geomagnetic sensor's susceptibility to interference from passing vehicles .

實施例四室內停車場管理系統 Embodiment 4 Indoor parking lot management system

若以第3圖為例，總共停車位為132個，Total Space：132；習知的技術需要約132/2=66個攝影機；本發明提出的方法，約只需要11-13個攝影機，加上132個車位感測器。加上配線，市電供應，以及通訊的IP需求。本發明的方法，所需成本至多是習知方法的一半甚至1/3。 Taking Figure 3 as an example, the total number of parking spaces is 132, Total Space: 132; the conventional technology requires approximately 132/2=66 cameras; the method proposed by the present invention only requires approximately 11-13 cameras, plus There are 132 parking sensors. Add wiring, power supply, and communication IP requirements. The cost of the method of the present invention is at most half or even 1/3 of the conventional method.

綜上所述，本發明的停車管理系統採用設置於一段車道旁的車位與設置於該車位的車位感測器，以及至少兩台相鄰的車牌號偵測器設置於該段車道的兩端，亦即第一車牌號偵測器與第二車牌號偵測器，藉由第一車牌號偵測器辨識進入該段車道的車輛車牌，第二車牌號偵測器辨識離開該段車道的車輛車牌，該車位感測器可即時精準偵測車輛佔位與否，並上傳佔位起迄時間的訊息給伺服器或雲端，可綜合研判停入車位的車輛車牌與車種。對於管理業者而言，其可減少甚至不需要巡車收費員，可大幅減少人事費用，即時精準有效管理毎一個車位，也可提供預約停車業務。對於車主而言，其可遠端連線查詢停車資訊，包括但不限於停車位使用狀態、停車時間、停車費用等，可自主線上報到與繳費，可以透過APP預約停車位、導航、與反向尋車。對於相關機關(例如交通管理單位)而言，也可運用此一資訊進行統計、稽查及管理。基此，相較於習知技藝而言，本發明係可提高街道上與街道外停車管理的便利性、時效性及正確性。 To sum up, the parking management system of the present invention adopts a parking space and a parking sensor installed in the parking space, and at least two adjacent license plate number detectors are installed at both ends of the road. , That is, the first license plate number detector and the second license plate number detector. The first license plate number detector recognizes the license plate of the vehicle entering the section of the lane, and the second license plate number detector recognizes the vehicle that leaves the section of the lane. Vehicle license plate, the parking sensor can accurately detect whether the vehicle is occupied or not, and upload the information of the start and end time of the vehicle to the server or the cloud, which can comprehensively determine the license plate and type of the vehicle parked in the parking space. For the management industry, it can reduce or even eliminate the need for patrol toll collectors, It can greatly reduce personnel costs, manage each parking space accurately and effectively in real time, and provide scheduled parking services. For car owners, they can remotely check parking information, including but not limited to parking space usage status, parking time, parking fees, etc., can report and pay online independently, and can reserve parking spaces, navigation, and reverse through APP Find a car. For related agencies (such as traffic management units), this information can also be used for statistics, inspections and management. Based on this, compared with conventional techniques, the present invention can improve the convenience, timeliness and accuracy of parking management on and off streets.

本發明，任何所屬技術領域中具有通常知識者，在不脫離本發明的精神和範圍內，當可作些許的更動與潤飾，故本發明的保護範圍當視後附的申請專利範圍所界定者為準。 The present invention, any person with ordinary knowledge in the technical field, can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be defined by the appended patent scope. Prevail.

21‧‧‧車位感測器 21‧‧‧Parking Sensor

23‧‧‧車牌號偵測器 23‧‧‧License plate number detector

Claims

一種停車管理系統，包含設置於停車場每一段車道旁的三個或三個以上的車位，與設置於該每一車位的車位感測器，設置於該段車道上空兩端的第一車牌號偵測器與第二車牌號偵測器，以及一伺服器：藉由第一車牌號偵測器辨識進入該段車道的車輛車牌號與時間，上傳至伺服器；第二車牌號偵測器辨識離開該段車道的車輛車牌號與時間，上傳至伺服器；另該車位感測器可即時精準偵測車輛停入與離開該車位的時間，並上傳車輛停入與離開該車位時間的訊息給伺服器，伺服器依照該車位被車輛停入與離開該車位的時間，搜尋第一與第二車牌號偵測器上傳的車牌號與時間，可綜合比對研判停入該車位的車輛車牌號及該車輛停入時間或離開該車位車牌號的車輛車牌號及該車輛離開時間。 A parking management system includes three or more parking spaces arranged beside each section of the lane in a parking lot, and a parking space sensor arranged in each section of the lane, and detecting the first license plate number arranged at both ends of the section of the lane And a second license plate number detector, and a server: the first license plate number detector recognizes the license plate number and time of the vehicle entering the lane, and uploads it to the server; the second license plate number detector recognizes and leaves The vehicle license plate number and time of the lane are uploaded to the server; in addition, the parking sensor can accurately detect the time when the vehicle enters and leaves the parking space, and uploads the information of the time when the vehicle enters and leaves the parking space to the server The server searches for the license plate number and time uploaded by the first and second license plate number detectors according to the time when the parking space is parked by the vehicle and leaves the parking space. It can comprehensively compare and judge the license plate number and The vehicle’s parking time or the vehicle’s license plate number and the vehicle’s departure time.

如申請專利範圍第1項的停車管理系統，其中的車牌號偵測器選自一可讀取車牌號的攝影機、可讀取車輛ID標籤(Tag)內車牌號的RFID讀取器、可讀取藍芽電子車牌內車牌號的藍芽讀取器、可讀取WIFI電子車牌內車牌號的WIFI讀取器、可讀取4G/5G電子車牌內車牌號的4G/5G讀取器。 For example, in the parking management system of item 1 of the scope of patent application, the license plate number detector is selected from a camera that can read the license plate number, an RFID reader that can read the license plate number in the vehicle ID tag (Tag), and readable Take the Bluetooth reader for the license plate number in the Bluetooth electronic license plate, the WIFI reader that can read the license plate number in the WIFI electronic license plate, and the 4G/5G reader that can read the license plate number in the 4G/5G electronic license plate.

如申請專利範圍第1項的停車管理系統，其中的車位感測器選自安裝於地表面之上或是與地面平或之下的地磁感測器、超音波感測器、紅外線感測器、微波感測器、雷達感測器，或其組合。 For example, in the parking management system of item 1 of the scope of patent application, the parking sensor is selected from geomagnetic sensors, ultrasonic sensors, infrared sensors installed on the ground surface, or level or below the ground , Microwave sensor, radar sensor, or a combination thereof.

如申請專利範圍第1項的停車管理系統，其中的車位感測器選自安裝於天花板的超音波感測器、紅外線感測器、微波感測器、雷達感測器，或其組合。 For example, in the parking management system of the first item in the scope of patent application, the parking sensor is selected from an ultrasonic sensor, an infrared sensor, a microwave sensor, a radar sensor, or a combination thereof installed on the ceiling.

如申請專利範圍第1項的停車管理系統，其中的車位感測器具備藍芽或WIFI或LPWA的無線通訊能力，能與車主的行動裝置上的藍芽或WIFI互通。 For example, in the parking management system of the first item of the patent application, the parking sensor has Bluetooth or WIFI or LPWA wireless communication capabilities, and can communicate with the Bluetooth or WIFI on the owner's mobile device.

如申請專利範圍第1項的停車管理系統，其中的車牌號偵測器進一步可安裝一藍芽作為定位用的信標(BEACON)，不斷廣播其所在地的經緯高度，此舉可讓車主的行動裝置APP能夠進行停車場的導航，並且能夠直接指派特定車位給車主。 For example, in the parking management system of item 1 of the scope of patent application, the license plate number detector can be further equipped with a bluetooth as a positioning beacon (BEACON), which continuously broadcasts the latitude and longitude of its location, which allows the owner to act The device APP can navigate the parking lot and can directly assign specific parking spaces to the car owner.

如申請專利範圍第1項的停車管理系統，其中的車牌號偵測器，具有藍芽或WIFI或LPWA或3G/4G/5G的通訊能力，能與車位感測器的藍芽或WIFI互通。 For example, in the parking management system of the first item of the patent application, the license plate number detector has Bluetooth or WIFI or LPWA or 3G/4G/5G communication capabilities, and can communicate with the Bluetooth or WIFI of the parking sensor.

如申請專利範圍第6項的停車管理系統，其中的行動裝置APP可提供以下至少一項服務：A.在室內獲得車牌號偵測器廣播的經緯高度，進行定位，獲得導航抵達指定停車格；B.車主只要按下車主APP上[自主報到]的按鍵，直接扣款，可享以分計費，打折付帳，無需去超商繳費，也無需下車在POS機繳費；C.可反向尋車，只要打開車主APP，一進入停車場，立即取得車牌號偵測器發送的經緯高度，手機即可定位自身，同時顯示停車場的地圖，以及愛車的位置，車主可接受導航，前往愛車的位置；D.可預約停車位，並透過導航，即可達成定時定點抵達預約的停車。 For example, the parking management system of item 6 of the scope of patent application, the mobile device APP can provide at least one of the following services: A. Obtain the latitude and longitude height broadcast by the license plate number detector indoors, perform positioning, and obtain navigation to reach the designated parking grid; B. The car owner only needs to press the button on the owner’s APP to directly deduct the payment, and he can enjoy split billing, discounted payment, no need to go to the supermarket to pay, and no need to get off the car to pay at the POS machine; C. Reverse To find a car, just open the owner’s APP, enter the parking lot, and immediately obtain the latitude and longitude height sent by the license plate number detector. The phone can locate itself and display the map of the parking lot and the location of the car. The owner can accept navigation and go to the location of the car. ; D. You can reserve a parking space, and through navigation, you can reach the scheduled parking at a fixed time.

一種停車管理方法，基於一個系統包含設置於停車場每一段車道旁的三個或三個以上的車位，與設置於該每一車位的車位感測器，設置於該段車道上空兩端的第一攝影機與第二攝影機，以及一伺服器；該方法包含下列步驟：步驟S41. 比較從第一攝影機進入的車牌號與從第二攝影機出去的車牌號；若某一車牌號有進入該車道的紀錄，但經過一個時間差，該車牌號未有離開該車道的紀錄，則將該車牌號的進入時間Tin，加以紀錄；該車牌號離開的時間Tout也加以記錄；若某一車牌號一進一出，而且相差時間在一個範圍內，則該車並未停入其間的車位，可加以忽略；步驟S42. 停車位上的車位感測器若被觸發確認有車停入，紀錄該觸發時間Tsin；步驟S43. 伺服器比較Tsin與Tin，若0<(Tsin-Tin)<Th則該新停入的車的車牌號，即為步驟S41於Tin取得的車牌號；如果只有Tin，沒有Tsin，則表示該車牌號違規停車，未停入車位內，或是仍未決定是否停入；步驟S44. 若車位已滿，並且如果只有Tin，沒有Tsin，則表示該車牌號違規停車，未停入車位內；步驟S45. 若是確認停車，則該車牌號，會經過一段時間，出現在車牌號離開的名單上，Tout；步驟S46. 停車格上的感測器若被觸發確認有車駛離，紀錄該觸發時間Tsout；步驟S47. 伺服器比較Tsout與Tout，若0<(Tout-Tsout)<Th則該新離開的車的車牌號，即為步驟S45於Tout取得的車牌號；如果只有Tout沒有Tsout，則表示該車牌號違規停車，未停入車格內。 A parking management method, based on a system including three or more parking spaces arranged beside each section of the lane in a parking lot, and a parking space sensor arranged in each parking space, which is installed in the section of the car The first camera and the second camera at both ends of the road, and a server; the method includes the following steps: step S41. Compare the license plate number entered from the first camera with the license plate number exited from the second camera; if a certain license plate number There is a record of entering the lane, but after a time difference, the license plate number does not have a record of leaving the lane, then the entry time Tin of the license plate number is recorded; the time Tout of the license plate number is also recorded; if a certain As soon as the license plate number is entered and the time difference is within a range, the car has not parked in the parking space in between and can be ignored; step S42. If the parking space sensor on the parking space is triggered to confirm that a car has parked, record The trigger time Tsin; step S43. The server compares Tsin and Tin, if 0<(Tsin-Tin)<Th, the license plate number of the newly parked car is the license plate number obtained in Tin in step S41; if there is only Tin If there is no Tsin, it means that the license plate number has illegally parked, has not parked in the parking space, or has not yet decided whether to park; step S44. If the parking space is full, and if there is only Tin and no Tsin, it means that the license plate number has illegally parked , Not parked in the parking space; Step S45. If parking is confirmed, the license plate number will appear on the list of license plate number departures after a period of time, Tout; Step S46. If the sensor on the parking grid is triggered, it is confirmed that there is When the car leaves, record the trigger time Tsout; Step S47. The server compares Tsout with Tout, if 0<(Tout-Tsout)<Th, the license plate number of the newly leaving car is the license plate number obtained at Tout in step S45 ; If there is only Tout and no Tsout, it means that the license plate number is illegally parked and not in the car compartment.

一種停車管理方法，基於一個系統包含設置於停車場每一段車道旁的三個或三個以上的車位，與設置於該每一車位的車位感測器，設置於該段車道上空兩端的第一攝影機與第二攝影機，以及一伺服器；該方法包含下列步驟：步驟S51. 第一攝影機紀錄毎一進入車輛的車牌號num與時間Tin，C(num,Tin)，並上傳至雲端；而第二攝影機紀錄毎一出去車輛的車牌號num與時間Tout，C(num,Tout)，並上傳至雲端；步驟S52. 停車位上的車位感測器若被觸發確認有車停入，紀錄該觸發時間Tsin，將該地點與Tsin上傳至雲端；步驟S53. 伺服器根據步驟S52取得信息的地點，找到對應的第一攝影機與第二攝影機的紀錄，並藉由步驟S52取得的Tsin，比較Tsin與Tin，若0<(Tsin-Tin)<Th則該新停入的車的車牌號，即為步驟1於Tin取得的車牌號；步驟S54. 若車位已滿，並且如果只有Tin，沒有Tsin，則表示該車牌違規停車，未停入車格內；步驟S55. 若是確認停車，則該車牌號，會經過一段時間，出現在車牌號離開的名單上，Tout；步驟S56. 停車格上的感測器若被觸發確認有車駛離，紀錄該觸發時間Tsout；步驟S57. 雲端比較Tsout與Tout，若0<(Tout-Tsout)<Th則該新離開的車的車牌號，即為步驟S55於Tout取得的車牌號；如果只有Tout，沒有Tsout，則表示該車牌號違規停車，未停入車格內。 A parking management method based on a system that includes three or more parking spaces arranged beside each section of the lane in a parking lot, a parking sensor set in each section of the lane, and first cameras arranged at both ends of the section of the lane And a second camera, and a server; the method includes the following steps: Step S51. The first camera records the license plate number num and time Tin, C (num, Tin) of each entering vehicle and uploads it to the cloud; and the second The camera records the license plate number num and time Tout, C (num, Tout) of each outgoing vehicle, and uploads it to the cloud; step S52. If the parking space sensor on the parking space is triggered to confirm that a car has entered, record the trigger time Tsin, upload the location and Tsin to the cloud; step S53. The server finds the corresponding records of the first camera and the second camera according to the location where the information is obtained in step S52, and compares Tsin and Tin based on the Tsin obtained in step S52 , If 0<(Tsin-Tin)<Th, the license plate number of the newly parked car is the license plate number obtained in Tin in step 1; step S54. If the parking space is full, and if there is only Tin and no Tsin, then It means that the license plate has parked illegally and has not been parked in the car compartment; step S55. If parking is confirmed, the license plate number will appear on the list of license plate number departures after a period of time, Tout; step S56. Sensing on the parking compartment If the device is triggered to confirm that there is a car leaving, record the trigger time Tsout; step S57. Compare Tsout with Tout in the cloud, if 0<(Tout-Tsout)<Th, then the license plate number of the newly leaving car, that is, in step S55 The license plate number obtained by Tout; if there is only Tout and no Tsout, it means that the license plate number is illegally parked and not in the car compartment.