200527328 玖、發明說明： 【發明所屬之技術領域】 本發明係關於一種多重扣款整合式多車道自由車流電子收 費系統，特別為一種執法單元與扣款單元獨立且具有電子收費 與門禁管理之功能，可結合車牌黑名單資料庫，做為贓車查緝 等應用之多重扣款整合式多車道自由車流電子收費系統。 【先前技術】 習用的電子收費系統，通常是以專用短程通信協議 (DSRC)、車輛定位系統(VPS)或微波單一扣款單元為電子收費系 統之扣款單元，並無法有效整合各種扣款單元的優點，舉例來 說，專用短程通信協議容易受到太陽及隔熱紙的影響，車輛定 位系統容易受到衛星定位不到的問題，微波容易受到外界干擾 及摻有金屬材質的隔熱紙影響。且當系統在轉換扣款單元時， 則執法單元與扣款單元的相關程式就必需相對配合更新。 由此可見，上述習用方式仍有缺失，實非一良善之設計者， 而亟待加以改良。 本案創作人鑑於上述習用所衍生的各項缺點，乃亟思加以 改進創新，並經多年苦心孤詣研究後，終於成功研發成本件多 重扣款機制之整合式多車道自由車流電子收費系統。 【發明内容】 本發明之目的之一在於提供一種具多重扣款機制之整合式 200527328 多車道自由車流電子收費系统，其可彈性使用不同扣款方式。 本發明之目的之二在於提供一種扣款單元與影像執法單元 獨立之多重扣款整合式多車道自由車流電子收費系統。 可達成上述發明目的之多重扣款整合式多車道自由車流電 子收費系統係由影像執法單元、扣款單元、後端處理單元所組 成。為有效整合各種扣款單元的優點與減少各種扣款單元的缺 點’扣款單兀可以採用車輛定位系統（VPS)、專用短程通信協議 (DSRC)或車輔&位系統加專用短程通信協議三 種方式。車輛定 位系統是由通用分組無線業務（GpRS)無線通訊單元、全球定位 系統（GPS)車輛定位單元、個人數位助理（pDA)車上單元所組 成。其扣款流程為事先預前定義扣款區域及執法區，再利用全 球定位系統車輛定位單元所擷取資料將全球定位系統結果與扣 款區比對’若於扣款區内，就會利用通用分組無線業務進行通 訊扣款並於執法區取像存證，若使用者未依規定扣款，則後端 匹配模組會加以整合判斷，做為逃費、違規告發之依據，並於 後端記錄查核。專用短程通信協議是由車上讀卡機（Vehicle200527328 发明 Description of the invention: [Technical field to which the invention belongs] The present invention relates to a multi-debit integrated multi-lane free traffic flow electronic toll collection system, particularly an independent law enforcement unit and debit unit with electronic toll collection and access control functions. It can be combined with the license plate blacklist database as an integrated multi-lane free-traffic electronic toll collection system for multiple deductions such as stolen vehicle detection. [Prior technology] The conventional electronic charging system usually uses a dedicated short-range communication protocol (DSRC), a vehicle positioning system (VPS), or a microwave single charge unit as the charge unit of the electronic charge system, and cannot effectively integrate various charge units The advantages are, for example, that the dedicated short-range communication protocol is easily affected by the sun and thermal insulation paper, the vehicle positioning system is vulnerable to satellite positioning problems, and the microwave is easily affected by external interference and thermal insulation paper mixed with metal materials. And when the system is converting the debit unit, the relevant programs of the law enforcement unit and the debit unit must be updated in coordination. It can be seen that the above-mentioned customary methods are still lacking. They are not a good designer and need to be improved. In view of the various shortcomings arising from the above-mentioned custom, the creator of this case was eager to improve and innovate. After years of painstaking research, he finally successfully developed an integrated multi-lane free-flow electronic toll collection system with a cost-reduction mechanism. [Summary of the Invention] One of the objectives of the present invention is to provide an integrated 200527328 multi-lane free traffic flow electronic toll collection system with multiple debit mechanisms, which can flexibly use different debit methods. Another object of the present invention is to provide an integrated multi-debit integrated multi-lane free traffic flow electronic toll collection system independent of a debit unit and an image enforcement unit. The multi-debit integrated multi-lane free-flow electronic toll collection system that can achieve the above-mentioned purpose of the invention is composed of an image law enforcement unit, a deduction unit, and a back-end processing unit. In order to effectively integrate the advantages of various debit units and reduce the disadvantages of various debit units, the debit unit can use a vehicle positioning system (VPS), a dedicated short-range communication protocol (DSRC), or a car auxiliary & bit system plus a dedicated short-range communication protocol. Three ways. The vehicle positioning system is composed of a general packet radio service (GpRS) wireless communication unit, a global positioning system (GPS) vehicle positioning unit, and a personal digital assistant (pDA) on-board unit. The deduction process is to define the deduction area and law enforcement area in advance, and then use the data captured by the GPS vehicle positioning unit to compare the GPS results with the deduction area. If it is in the deduction area, it will be used The general packet wireless service conducts communication deductions and obtains image deposit certificates in the law enforcement area. If the user does not deduct according to the regulations, the back-end matching module will integrate the judgments as the basis for evading charges and reporting violations, and later End record check. The dedicated short-range communication protocol is implemented by the vehicle's card reader (Vehicle
Control Unit ’ VCU)與車道系統(Road Top Unit，RTU)所組成， 會同時驅動控制卡板，而控制卡板便會設定車道系統（RTu)，以 做通訊之準備。影像執法單元由取像單元、車牌辨識單元、接 圖（Image Registration)單元及車型辨識單元所組成。取像單元包 3刖照（後照）感應早元、刖照（後照）光麵合（ecu)攝影機、影像 200527328 擷取卡（Frame Grabber)、數位輸出入卡（Digital Input/Output)等 所組成。在程式的執行過程中，會以問卷的方式每間隔一定時 間的掃瞄週期去檢查控制卡板是否有資料進來，而控制卡板會 以每間隔一定時間的掃瞄週期去檢查車道系統是否有車上讀卡 機進入通訊區。當有車輛進入前照（後照）的感應區域且車上有車 上讀卡機單元，則扣款單元會執行扣款動作。 當車輛定位系統與專用短程通信協議整合使用時，必須事 先預前定義扣款區域及執法區，且當車輛進入扣款區時，如果 全球定位系統因衛星定位失敗而失效時，可以專用短程通信協 議來輔助；同理，如果專用短程通信協議因太陽光或隔熱紙的 影響而失效時，則可以車輛定位系統來輔助；如果車輛定位系 統與專用短程通信協議同時動作時，則後端可以利用匹配來整 合扣款資料；如果車輛定位系統及專用短程通信協議同時失效 時，則也還有執法系統當作最後的防線。 此外，執法單元與扣款單元是分別獨立的個體，所有的匹 配動作都是在後端進行，執法單元只是提供車牌影像、辨識結 果及通行紀錄給後端，而扣款單元只是提供扣款紀錄給後端， 所以如果扣款單元損壞時，並不會造成執法單元無法執行，同 理如果執法單元損壞時，並不會造成扣款單元無法執行，所以 執法系統具有很強的獨立性與穩定性，也因此不論扣款系統採 用那一種扣款方式，執法系統都不必重新修改影像執法單元的 12 200527328 程式，具有很大的彈性。 【實施方式】 請參閱圖一所示，係本發明所提供之系統架構流程圖。由 於影像執法單元1與扣款單元2(由車輛定位系統及專用短程通 信協議組成）是分別獨立的個體，所有的匹配動作都是在後端處 理單元3(由電腦自動匹配單元31、人工比對單元32及罰單寄 發機制33組成）進行，影像執法單元1只是提供車牌影像、辨 識結果及通行紀錄給後端3 ’而扣款單元2只是提供扣款紀錄給 後端3，所以如果扣款單元2損壞時，並不會造成執法單元1 無法執行，同理如果執法單元1損壞時，並不會造成扣款單元2 無法執行，所以執法系統具有很強的獨立性與穩定性，也因此 不論扣款系統採用那一種扣款方式，執法系統都不必重新修改 影像執法單元1的程式，具有很大的彈性。 由於取像單元11藉由前（後）照感應單元12觸發取像，所以 可以避免因不同的車速而導致的取像失敗或因大小車混合所導 致的取像失敗。此外，為避免產生同一台車多次取像導致取像 資料過多而造成後端比對機制的困擾，所以我們採用鎖定（lock) 的方法來解決此一問題。所謂鎖定是指當前（後）照感應單元12 感應到車輛進入感應區後自動取像，並鎖定住取像單元，直到 前（後）照感應單元12感應車輛離開感應區。另可增設一黑名單 管理單元4,係利用影像執法單元所辨識出來的車牌號碼與扣款 13 200527328 單元2所辨識出來的扣款單元ID來查詢黑名單資料庫41，將 來自黑名單資料庫41之資料匯入電腦自動匹配單元31進行整 合，若查詢出來是黑名單則交由警察單位處理。 請參閱圖二，為本發明之系統資料流程圖。由於是採用登 記制，在正常使用下，扣款單元ID與車牌號碼在資料庫當中 是以一對一方式對應的，當車牌辨識5後，若自動辨識成功51， 則判斷影像中車號是否有登記52，結果區分成二大類，亦即影 像中車號係無登記者或影像中車號係有登記者，且若車牌辨識 成功，則寄發罰單。無登記者原則上不扣款521，但若是因為車 號無法自動辨識者51，還必需經由後端人工比對機制53處理， 才不致於有誤罰的可能性。而有登記者則可與扣款系統的扣款 單元識別碼（ID)交由後端人工處理機制53進行自動比對。如果 後端人工比對機制53人工辨識的結果發現車牌取像成功54,則 又可依扣款成功與否55區分成二大類，亦即車牌取像成功且扣 款成功56及車牌取像成功但扣款失敗57二大類。如果沒有取 像資料，則又可依扣款情形區分成二大類，亦即車牌取像失敗 但扣款成功58及車牌取像失敗且扣款失敗59二大類。當然， 只有車牌取像成功但扣款失敗才需寄發罰單571。當然，也就是 因為我們採用登記制，所以不會有因扣款單元號碼與車牌號碼 匹配錯而誤罰。 請參閱圖三，為本發明之四車道的光耦合攝影機架設位置 14 200527328 圖。如圖所示，我們分別將上行的前照光耦合攝影機命名為 UF1〜5，上行後照的光耦合攝影機命名為UB1〜5，下行的前照 光耦合攝影機命名為DF1〜5，下行的後照光耦合攝影機命名為 DB1〜5 〇在正常行駛的情況下，如果上行車輛感應區4(6)觸發， 則UF2(UF4)會進行取像；如果是下行車輛感應區丨（3)觸發，則 DF2(DF4)會進行取像。當車輛變換車道時，如果使上行車輛感 應區5觸發，則UF1、UF3、UF5會進行取像；如果使下行車輛 感應區2觸發，則DF1、DF3、DF5會進行取像。為避免徒增後 端人工處理機制的困擾，如果車牌辨識程式能成功辨識UF3或 DF3的車牌影像，則UF1、UF5或DF1、dF5並不會使用；相 對地，如果車牌辨識程式無法成功辨識UF3或DF3的車牌影 像，則會自動對UF卜肥或DF1、DF5進行車牌辨識，並: 結果送交後端人工比對機制進行自動比對。 由於多車道自由車流可能會同時摻雜大貨車、小轉車，則 小轎車的車牌很有可能被大貨車遮住，所以，我們提出使用後 照的方式來加以補強。亦即，當備測到平均車速過慢，則自動 啟動後照功能，換言之’在塞車的情况下，如果上行車輛感應 區丄觸發，則UF2、脈會進行取像；如果上行車輛感應區3 觸發’則UF4、UB4會進行取像；如果是下行車㈣應區㈣ 發，則DF2、DB2會進行取像；如果是下行車輛感應區6觸發， 則DF4 ' DB4會進行取像·，當車輛變換車道時，如果使上行車 15 200527328 輛感應區2觸發，貝ij UFl、UF3、UF5、UBl、UB3、UB5會進 行取像；如果使下行車輛感應區5觸發，則DF1、DF3、DF5、 DB卜DB3、DB5會進行取像。為避免徒增後端人工處理機制的 困擾’如果車牌辨識程式能成功辨識UF3或DF3的車牌影像， 則 UF1、UF5、UB1、UB5 或 DF1、DF5、UB1、UB5 並不會使 用；相對地，如果車牌辨識程式無法成功辨識UF3、UB3或DF3、 DB3的車牌影像，貝會自動對UF1、UF5、UB1、UB5或DF1、 DF5、DB1、DB5進行車牌辨識，並將結果送交後端人工處理機 制進行自動比對。此一方法形同第一位專家先利用UF2、UF3、 UF4或DF2、DF3、DF4的前照影像來進行檢驗車牌取像是否成 功，如果不成功，才動用第二位專家針對UF1、UF5或DF1、 DF5的前側照影像進行檢驗，如果還不成功，則動用第三位專 家針對UB2、UB3、UB4或DB2、DB3、DB4的後照影像進行 檢驗車牌取像是否成功’如果不成功，再動用第四位專家針對 的後侧照影像UB1、UB5或DB1、DB5進行檢驗，最後一道關 卡’則是操作員這個第五位專家來進行人工比對。運用這種多 重專家的方式可有效減少取像失敗的機率，大幅提昇電子收費 系統的可靠度。雖然會增加取像資料量，但利用多重專家的方 式可加速系統運算時間。使用前後照搭配的方式，可有效解決 塞車時大貨車遮蔽小轎車的問題，也可避免為規避繳費而將車 距拉小所導致之取像失敗。 200527328 在本發明當中，扣款單元可以採用車㈣位系統、專用短 程通信協議或車輛定位系統加專用短程通信協議三種方式。車 辅定位系統是由通时組無線轉無線龍單元、全球定位系 統車蚊位單元、個人數位助理(PDA)車上單元所組成。其扣款 流㈣事^預前^義扣款區域及執法區，再利用全球定位系統 車輛疋位早儿所擷取資料將全球定位系統結果與扣款區比對， 若於扣款區内’就會利用通用分組無線業務進行通訊扣款並於 執会區取像存迅，趣用者未依規定扣款，職端匹配模組會 加以整合判斷’做為逃費、違規告發之依據，並於後端記錄查 核。專驗料㈣議是由車上讀卡機與車料統所組成，會 同時驅動㈣卡板，㈣制卡板便會設定車道系統，以做通訊 之準備。在程式的執行過財，會以輪詢时式每間隔一定時 _掃晦·去檢查控制卡板是㈣資料進來，而控制卡板會 以母間隔-定時間的掃猫週期去檢查車道系統是否有車上讀卡 機進入通訊區。當有車輛進入前照(後照)的感應區域且車上有車The Control Unit ’VCU) and the Road Top Unit (RTU) will drive the control card at the same time, and the control card will set the lane system (RTu) to prepare for communication. The image law enforcement unit consists of an image acquisition unit, a license plate recognition unit, an image registration unit, and a vehicle identification unit. Image acquisition unit package 3 photo (backlight) sensing early, photo (backlight) smooth surface (ecu) camera, image 200527328 frame grabber, digital input / output, etc. Composed of. During the execution of the program, a questionnaire is used to check the control card board for data at regular intervals, and the control card board is used to check the lane system for scanning data at regular intervals. The card reader in the car enters the communication area. When a vehicle enters the sensing area of the front (backlight) and there is a card reader unit on the vehicle, the deduction unit will perform the deduction action. When the vehicle positioning system is integrated with the dedicated short-range communication protocol, the deduction area and law enforcement area must be defined in advance, and when the vehicle enters the deduction area, if the global positioning system fails due to satellite positioning failure, dedicated short-range communication can be used Similarly, if the dedicated short-range communication protocol fails due to the influence of sunlight or thermal insulation paper, the vehicle positioning system can assist; if the vehicle positioning system and the dedicated short-range communication protocol operate at the same time, the back end can Use matching to integrate charge data; if the vehicle positioning system and the dedicated short-range communication protocol fail at the same time, there is also a law enforcement system as the last line of defense. In addition, the law enforcement unit and the deduction unit are independent individuals. All matching actions are performed on the back end. The law enforcement unit only provides license plate images, identification results and traffic records to the back end, while the deduction unit only provides deduction records. To the back end, so if the deduction unit is damaged, it will not cause the law enforcement unit to execute. Similarly, if the law enforcement unit is damaged, it will not cause the deduction unit to execute, so the law enforcement system has strong independence and stability. Therefore, no matter which type of deduction method is used by the deduction system, the law enforcement system does not need to re-modify the 12 200527328 program of the image law enforcement unit, which has great flexibility. [Embodiment] Please refer to FIG. 1, which is a flowchart of a system architecture provided by the present invention. Since the image law enforcement unit 1 and the deduction unit 2 (composed of the vehicle positioning system and the dedicated short-range communication protocol) are separate entities, all matching actions are performed in the back-end processing unit 3 (by the computer automatic matching unit 31, manual comparison It is composed of the unit 32 and the ticket issuing mechanism 33. The image law enforcement unit 1 only provides the license plate image, the recognition result and the passing record to the back end 3 ', while the deduction unit 2 only provides the deduction record to the back end 3, so if the When the damage to the law enforcement unit 2 is not caused, the law enforcement unit 1 cannot be executed. Similarly, if the law enforcement unit 1 is damaged, the deduction unit 2 cannot be executed. Therefore, the law enforcement system has strong independence and stability. Therefore, no matter which type of deduction method is used by the deduction system, the law enforcement system does not need to re-modify the program of the image law enforcement unit 1, which has great flexibility. Since the image capturing unit 11 triggers the image capturing by the front (back) photo sensing unit 12, it is possible to avoid image capturing failure due to different vehicle speeds or image capturing failure due to mixing of large and small vehicles. In addition, in order to avoid the trouble of back-end comparison mechanism caused by too many acquisition data caused by multiple acquisitions of the same vehicle, we adopt a lock method to solve this problem. The so-called lock means that the current (back) photo sensing unit 12 automatically captures an image when the vehicle enters the sensing area, and locks the image capturing unit until the front (back) photo sensing unit 12 senses that the vehicle leaves the sensing area. In addition, a blacklist management unit 4 can be added, which uses the license plate number and deduction identified by the image enforcement unit 13 200527328 The deduction unit ID identified by unit 2 to query the blacklist database 41, which will come from the blacklist database The data of 41 is imported into the computer automatic matching unit 31 for integration. If the query is a blacklist, it is handed over to the police unit for processing. Please refer to FIG. 2 for a system data flow chart of the present invention. Due to the registration system, under normal use, the deduction unit ID and the license plate number correspond one-to-one in the database. After the license plate recognition is 5, if the automatic recognition is successful, 51, determine whether the vehicle number in the image is There are registrations 52, and the results are divided into two categories, that is, those who have no registration in the image or those who have registration in the image, and if the license plate is successfully identified, a ticket is issued. Those who do not register will not be charged 521 in principle. However, if the car number cannot be identified automatically by the person 51, it must also be processed by the back-end manual comparison mechanism 53 to avoid the possibility of mistaken punishment. Those who have registered can hand over the debit unit identification code (ID) of the deduction system to the back-end manual processing mechanism 53 for automatic comparison. If the result of manual recognition of the back-end artificial comparison mechanism 53 finds that the license plate is successfully acquired 54, it can be divided into two categories according to the success or failure of the deduction 55, that is, the license plate is successfully acquired and the charge is successfully 56 and the license plate is successfully acquired. But the chargeback failed 57 two major categories. If there is no image acquisition data, it can be divided into two categories according to the deduction situation, that is, the license plate acquisition fails but the deduction is successful58 and the license plate acquisition fails and the deduction fails59. Of course, only if the license plate is successfully acquired but the deduction fails, a ticket 571 needs to be sent. Of course, because we use the registration system, there will be no penalty due to the mismatch between the debit unit number and the license plate number. Please refer to FIG. 3, which is a diagram showing the setting position of the light-coupling photography rack for the four lanes of the present invention. As shown in the figure, we named the upstream front-coupled camera as UF1 ~ 5, the upstream back-coupled camera as UB1 ~ 5, the downstream front-coupled camera as DF1 ~ 5, and the downstream back-coupled camera. The camera is named DB1 ~ 5. Under normal driving conditions, if the upward sensing area 4 (6) is triggered, then UF2 (UF4) will take an image; if the downward sensing area 丨 (3) is triggered, then DF2 ( DF4) will acquire images. UF1, UF3, and UF5 will acquire images when the upshift vehicle sensing area 5 is triggered when the vehicle changes lanes; if DF1, DF3, DF5 will be acquired if the downshift vehicle sensing area 2 is triggered. In order to avoid the trouble of increasing the manual processing mechanism at the back end, if the license plate recognition program can successfully recognize the license plate image of UF3 or DF3, UF1, UF5 or DF1, dF5 will not be used; on the contrary, if the license plate recognition program cannot successfully recognize UF3 Or DF3 license plate image, it will automatically recognize the license plate of UF fertilizer or DF1, DF5, and: The result will be sent to the back-end manual comparison mechanism for automatic comparison. Since the multi-lane free traffic may be mixed with large trucks and small transfers at the same time, the license plate of a small car is likely to be covered by the large trucks. Therefore, we propose to use the backlight method to reinforce it. That is, when the average speed of the vehicle is too slow, the backlight function is automatically activated. In other words, in the case of a traffic jam, if the sensing area of the upward vehicle is triggered, the UF2 and the pulse will be captured; if the upward vehicle sensing area is 3 Trigger ', then UF4, UB4 will take the image; if it is a downward vehicle response zone, then DF2, DB2 will take an image; if it is a downward vehicle sensor zone 6 trigger, then DF4' DB4 will take an image, when When the vehicle changes lanes, if the up-going 15 200527328 sensing area 2 is triggered, the beijing UF1, UF3, UF5, UB1, UB3, UB5 will take images; if the downward sensing area 5 is triggered, DF1, DF3, DF5 DB, DB3, DB5 will take images. In order to avoid the trouble of increasing the back-end manual processing mechanism, 'If the license plate recognition program can successfully recognize the license plate image of UF3 or DF3, then UF1, UF5, UB1, UB5 or DF1, DF5, UB1, UB5 will not be used; relatively, If the license plate recognition program cannot successfully recognize the license plate image of UF3, UB3 or DF3, DB3, Bay will automatically recognize the license plate of UF1, UF5, UB1, UB5 or DF1, DF5, DB1, DB5, and send the result to the back end for manual processing. Mechanism for automatic comparison. This method is similar to the first expert who uses the front image of UF2, UF3, UF4 or DF2, DF3, DF4 to check whether the license plate acquisition is successful. If it is unsuccessful, then use the second expert to target UF1, UF5 or DF1, DF5's front side view images are checked. If it is not successful, use a third expert to check whether the license plate image acquisition is successful for UB2, UB3, UB4 or DB2, DB3, DB4. If it is not successful, then The fourth expert used the backside image UB1, UB5 or DB1, DB5 for inspection, and the last level was the fifth expert, the operator, for manual comparison. The use of this multi-expert approach can effectively reduce the chance of image acquisition failure and greatly improve the reliability of the electronic toll collection system. Although it will increase the amount of image data, using multiple experts can speed up the system operation time. The combination of front and rear photos can effectively solve the problem of large trucks covering cars when there is a traffic jam, and it can also avoid the failure of image acquisition caused by narrowing the distance between vehicles to avoid payment of fees. 200527328 In the present invention, the debit unit can adopt three modes: a car seat system, a dedicated short-range communication protocol, or a vehicle positioning system plus a dedicated short-range communication protocol. The vehicle-assisted positioning system is composed of the Tongshi group wireless to wireless dragon unit, the global positioning system car mosquito bit unit, and the personal digital assistant (PDA) on-board unit. Its debit flow is ^ forecast ^ meaning the deduction area and law enforcement area, and then use the data captured by the GPS vehicle niches early to compare the GPS results with the deduction area. 'It will use the general packet wireless service for communication deductions and take images from the conference area. If the fun users do not deduct according to regulations, the job matching module will integrate the judgment' as the basis for fee evasion and violation reporting. , And check back-end records. The special material inspection proposal is composed of the on-board card reader and the vehicle material system, which will drive the card board at the same time, and the card board will set the lane system to prepare for communication. After the program has been rich, it will use the polling method every interval to _ scan the dark to check the control card is to enter the data, and the control card will check the lane system at the interval between the female and the fixed time. Is there a card reader in the car entering the communication area. When a vehicle enters the sensing area of the front light (back light) and there is a car on the car
上讀卡機單元，則扣款單元會執行扣款動作。當車輛定位系統 與專用短程通信協議整合使用時，必須事_前定義扣款區域 及執法區，且當車輛進入扣款區時，如果全球定位系統因衛星 定位失敗而失效時，可以專驗程通信協議來輔助；同理如 果專用短程通信協議因太陽光或隔熱紙的影響而失效時，則可 以車描定位系統來辅助；如果車輛定位㈣與專用短程通信協 17 200527328 議同時動作時，則後端可以利用匹配來整合扣款資料；如果車 輛定位系統及專用短程通信協議同時失效時，則也還有執法系 統當作最後的防線。 如圖四〜七所示，後端人工比對機制由車柄通行資料自動 匹配單元與人工操作介面單元所組成。車輛通行資料自動匹配 單元會先將扣款與執法系統匯入資料庫對應表格中，然後第一 階段依照本身車道與跨車道之順序，在一定的時間區間之内， 將扣款之扣款單元登記車號與執法辨識車號進行比對，若完全 相同者，紀錄彼此之匹配關係。第二階段依照本身車道與跨車 道之順序，在一定的時間區間之内，將扣款之扣款單元登記車 號與執法辨識車號進行比對，若四個（含）以上車牌號碼以上相同 者，紀錄彼此之匹配關係。若第一、二階段完成後，仍然沒有 完成匹配關係，則依照本身車道與跨車道之順序，在一定的時 間區間之内，利用扣款之時間與執法時間計算時間差（計算至微 秒），將時間差最接近者，顯示在螢幕上供操作者進行人工比 對。若最後有車牌影像但沒有扣款單元ID則逕行告發。由於傳 統的執法系統，為了取得完整的全景，所以分成多台攝影機分 別取像，但是如果後端匹配失敗，則操作員必須從大量的圖形 檔當中找出關聯性，但比較沒有連續空間的感覺。為解決此一 問題提出使用接圖（Image Registration)的技術將同一時間多台 攝影機之個別影像加以整合成環場影像，讓操作員在進行人工 200527328 比對的時候，能夠比較有連續空間的感覺，也可以簡化操作員 人工比對的時間。 此外，為因應不同車型使用不同的價錢，本發明提出利用 雷射感測器進行車型辨識，利用雷射感測器所連續取得之輪廓 (profile)資訊加以判斷車型，如果系統應用在電子收費系統，則 系統可依據影像執法系統所判斷出來之車輛車型與車上讀卡機 所設定的車型來做判斷，如果不同，則影像執法系統就可以逕 行開立罰單。 後端人工比對機制使用者介面單元在本發明當中是以網頁 的方式呈現。人工操作介面可分為單筆式及多筆式。單筆式一 次處理一筆匹配資料，在左方顯示單筆扣款資料（包含車牌辨識 單元所辨識出來的車牌號碼、扣款單元所取得的扣款單元識別 碼（ID)、與扣款單元識別碼匹配的車牌號碼、取像時間、扣款時 間、曾經經過的車道、平均車速、黑名單狀態等），右方列出可 能匹配之執法資料，在可能的匹配取像圖片上加上可能性的百 分比，以利操作員進行判斷。並提供核取方塊以人工勾選的方 式勾選匹配資料（可複選），完成後按鈕進行下一筆之匹配動作。 由於可能同一台車多次取像，所以後端處理機制採用可複選的 方式將取像資料進行整合，以減少不明影像的張數。此外，操 作員可以藉由人眼辨識車牌號碼與網頁所提供之資料來判斷那 幾張取像資料應該與扣款資料進行整合。但是由於隔熱紙所造 19 200527328 成的紅外線折射、反射、繞射或網路傳輸延遲或雜訊所造成的 延遲等等因素所導致的執法時間順序未必等於扣款時間順序。 因此如果單純使用時間來當做比對的重要因素可能會造成比對 錯誤。因此本發明提供了另外一種模式，也就是多筆式。多筆 式一次可同時看到在一時間區間内的所有扣款資料及取像資 料，操作者可以以比較宏觀的角度來進行人工比對的動作。同 樣地，在可能的匹配取像圖片上加上可能性的百分比，以利操 作員進行判斷。 不論是單筆式或多筆式使用者操作介面當中的可能性分 析，均係以車牌辨識單元所辨識出來的車牌號碼、與扣款單元 識別碼匹配的車牌號碼、車輛的顏色、車型資料、取像時間順 序、扣款時間順序及曾經經過的車道等因素配合權重的觀念來 加以計算可能性的百分比。 本發明與其它習用技術相互比較時，更具有下列之優點： 先前的影像執法系統，通常是專用短程通信協議、車輛定位系 統或微波系統分別與影像執法程式配合使用，並不具有整合 性。且當系統在轉換扣款單元時，執法端的程式就必需更新。 由於執法單元與扣款單元是分別獨立的個體，所有的匹配動作 都是在後端進行，執法單元只是提供車牌影像、辨識結果及通 行紀錄給後端，而扣款單元只是提供扣款紀錄給後端，所以如 果扣款單元損壞時，並不會造成執法單元無法執行，同理如果 200527328 執法單元損壞時’並不會造成扣款單元無法執行，所以執法系 統具有很強的獨立性與穩定性，也因此不論扣款系統採用那— 種扣款方式，執法系統都不必重新修改影像執法單元的程式， 具有很大的彈性。 由於專用短程通信協議容易受到太陽及隔熱紙的影響，車 輛定位系統容易受到衛星定位不到的問題，微波容易受到外界 干擾及摻有金屬材質的隔熱紙的影響。當車輛定位系統盘專用 短程通信協議整合使用時，必須事先預前定義扣款區域及執法籲 區’且當車輛進入扣款區時’如果全球定位系統因衛星定位失 敗而失效時，可以專用短程通信協議來輔助；同理，如果專用 紐程通信協議因太陽光或隔熱紙的影響而失效時，則可以車輛 定位系統來輔助；如果車輛定位系統與專用短程通信協議同時 動作時，則後端可以利用匹配來整合扣款資料；如果車輛定位 系統及專用短程通信協議同時失效時，則也還有執法系統當作 最後的防線；也由於取像單元與扣款單元是由後端進行匹配， · 所以執法單元與扣款單元是具有獨立性的，且執法單法具有彈 性去接受各種不同的扣款單元。 上列詳細說明乃針對本發明之一可行實施例進行具體說 明’惟該實施例並非用以限制本發明之專利範圍，凡未脫離本 發明技藝精神所為之等效實施或變更，均應包含於本案之專利 範圍中。 21 200527328 綜上所述，本案不僅於技術思想上確屬創新，並具備習用 方法所不及之上述多項功效，已充分符合新穎性及進步性發明 專利要件，爰依法提出申請，懇請貴局核准本件發明專利申 請案，以勵發明，至感德便。 【圖式簡單說明】 請參閱以下有關本發明之附圖，將可進一步瞭解本發明之 技術内容及其目的功效；有關該實施例之附圖為： 圖一為本發明之系統架構流程圖； 圖二為本發明之系統資料流程圖； 圖三為本發明光耦合攝影機之架設位置圖； 圖四為本發明之後端人工比對機制第一階段流程圖； 圖五為本發明之後端人工比對機制第二階段流程； 圖六為本發明之後端人工比對機制第三階段流程； 圖七為本發明之後端人工比對機制第四階段流程圖。 【主要部分代表符號】 1 影像執法單元 11 取像單元 12 前（後）照感應單元 2 扣款單元 3 後端處理單元 31 電腦自動匹配單元 22 200527328 32 33 4 41 人工比對操作單元 罰單寄發單元 黑名單管理單元 黑名單資料庫If the card reader unit is connected, the debit unit will perform the debit action. When the vehicle positioning system is integrated with the dedicated short-range communication protocol, the deduction area and law enforcement area must be defined in advance, and when the vehicle enters the deduction area, if the global positioning system fails due to satellite positioning failure, it can be specially inspected. Communication protocol to assist; Similarly, if the special short-range communication protocol fails due to the influence of sunlight or thermal insulation paper, you can use the car mapping positioning system to assist; if the vehicle positioning ㈣ and the special short-range communication protocol 17 200527328 Then the back end can use matching to integrate the charge data; if the vehicle positioning system and the dedicated short-range communication protocol fail at the same time, there is also a law enforcement system as the last line of defense. As shown in Figures 4-7, the back-end manual comparison mechanism is composed of an automatic matching unit for handle data and a manual interface unit. The vehicle traffic data automatic matching unit will first import the deduction and law enforcement system into the database corresponding table, and then in the first stage, according to the order of its own lane and cross-lane, the deduction unit will be deducted within a certain period of time. The registered car number is compared with the law enforcement identification car number. If they are exactly the same, the matching relationship between them is recorded. In the second stage, according to the order of its own lane and cross lanes, within a certain time interval, the deduction unit registration vehicle number is compared with the law enforcement identification vehicle number. If four or more license plate numbers are the same Or, record the matching relationship with each other. If the matching relationship is still not completed after the first and second phases are completed, the time difference between the time of the deduction and the law enforcement time is calculated within a certain period of time in accordance with the order of its own lane and cross lanes (calculated to microseconds). The closest time difference is displayed on the screen for manual comparison by the operator. If there is a license plate image but no debit unit ID at the end, go ahead and report it. Due to the traditional law enforcement system, in order to obtain a complete panorama, it is divided into multiple cameras to take images separately, but if the back-end matching fails, the operator must find the correlation from a large number of graphics files, but there is no sense of continuous space. . In order to solve this problem, it is proposed to use the technology of Image Registration to integrate the individual images of multiple cameras at the same time into a ring field image, so that the operator can have a sense of continuous space when performing manual 200527328 comparison. , Can also simplify the time of manual comparison of operators. In addition, in order to use different prices for different vehicle models, the present invention proposes to use a laser sensor to identify the vehicle model, and to use the profile information continuously obtained by the laser sensor to determine the vehicle model. If the system is applied to an electronic toll collection system , The system can make a judgment based on the vehicle model determined by the image enforcement system and the model set by the card reader on the car. If they are different, the image enforcement system can issue a ticket. The user interface unit of the back-end manual comparison mechanism is presented in the form of a web page in the present invention. Manual operation interface can be divided into single pen type and multi-pen type. One-shot processing of one piece of matching data at a time, displaying single charge data (including the license plate number identified by the license plate recognition unit, the deduction unit identification code (ID) obtained by the deduction unit, and the deduction unit identification on the left) Code matching license plate number, image acquisition time, deduction time, lanes passed, average speed, blacklist status, etc.), the law enforcement information of possible matches is listed on the right, and possibility is added to the pictures of possible matches Percentage for operator judgment. It also provides a check box to manually select the matching data (can be selected), and then press the button to perform the next matching action. Since the same vehicle may take images multiple times, the back-end processing mechanism integrates the acquisition data in a selectable manner to reduce the number of unknown images. In addition, the operator can judge by human eyes the license plate number and the information provided on the webpage to determine which image acquisition data should be integrated with the deduction data. However, due to factors such as infrared refraction, reflection, diffraction, or network transmission delay or delay caused by noise caused by thermal insulation paper 19 200527328, the law enforcement time sequence may not be equal to the deduction time sequence. Therefore, if you simply use time as an important factor of comparison, it may cause comparison errors. Therefore, the present invention provides another mode, which is a multi-stroke mode. With multiple strokes, you can see all the deduction data and image acquisition data in a time interval at the same time. The operator can perform the manual comparison operation from a relatively macro perspective. Similarly, the percentage of likelihood is added to the possible matching image to facilitate the operator's judgment. Regardless of the possibility analysis in the single-pen or multi-pen user interface, the license plate number identified by the license plate recognition unit, the license plate number matched with the deduction unit identification code, the color of the vehicle, the model information, Factors such as chronological order, deductive chronological order, and lanes used to calculate the percentage of probability are combined with the concept of weight. The present invention has the following advantages when compared with other conventional technologies: The previous image law enforcement systems, which are usually dedicated short-range communication protocols, vehicle positioning systems, or microwave systems, are used in conjunction with image law enforcement programs, and are not integrated. And when the system is converting the debit unit, the law enforcement program must be updated. Since the law enforcement unit and the deduction unit are separate entities, all matching actions are performed on the back end. The law enforcement unit only provides license plate images, identification results and traffic records to the back end, while the deduction unit only provides deduction records to Backend, so if the deduction unit is damaged, it will not cause the law enforcement unit to execute. Similarly, if 200527328 is damaged, the law enforcement unit will not cause the deduction unit to be unenforceable, so the law enforcement system has strong independence and stability. Therefore, no matter what kind of deduction method the deduction system uses, the law enforcement system does not need to re-modify the program of the image law enforcement unit, which has great flexibility. Because the dedicated short-range communication protocol is easily affected by the sun and thermal insulation paper, the vehicle positioning system is vulnerable to satellite positioning problems, and the microwave is easily affected by external interference and thermal insulation paper mixed with metal materials. When the vehicle positioning system disk-specific short-range communication protocol is integrated and used, the deduction area and law enforcement appeal area must be defined in advance, and when the vehicle enters the deduction area ', if the global positioning system fails due to satellite positioning failure, the short-range can be used exclusively. The communication protocol is used to assist; Similarly, if the dedicated New York communication protocol fails due to the influence of sunlight or thermal insulation paper, the vehicle positioning system can be used to assist; if the vehicle positioning system and the dedicated short-range communication protocol operate at the same time, then the latter The end can use matching to integrate deduction data; if the vehicle positioning system and the dedicated short-range communication protocol fail at the same time, there is also a law enforcement system as the last line of defense; and because the imaging unit and the deduction unit are matched by the back end Therefore, the law enforcement unit and the deduction unit are independent, and the law enforcement single law has the flexibility to accept a variety of different deduction units. The above detailed description is a specific description of one of the feasible embodiments of the present invention. 'However, this embodiment is not intended to limit the patent scope of the present invention. Any equivalent implementation or change without departing from the technical spirit of the present invention should be included in Within the scope of the patent in this case. 21 200527328 In summary, this case is not only innovative in terms of technical ideas, but also has the above-mentioned multiple effects that are not used by conventional methods. It has fully met the requirements for novel and progressive invention patents, and applied for it in accordance with the law. Application for invention patents, to encourage invention, to the utmost. [Brief description of the drawings] Please refer to the following drawings of the present invention to further understand the technical content of the present invention and its purpose and effect. The drawings related to this embodiment are: Figure 1 is a flowchart of the system architecture of the present invention; Figure 2 is a flowchart of the system data of the present invention; Figure 3 is a diagram of the installation position of the optical coupling camera of the present invention; Figure 4 is a first stage flowchart of the rear-end manual comparison mechanism of the present invention; Figure 5 is a rear-end manual comparison of the present invention The second stage flow of the mechanism; Figure 6 is the third stage flow of the rear end manual comparison mechanism of the present invention; and Figure 7 is the fourth stage flow chart of the back end manual comparison mechanism of the present invention. [Representative symbols of main parts] 1 Image law enforcement unit 11 Image acquisition unit 12 Front (back) photo sensing unit 2 Deduction unit 3 Back-end processing unit 31 Computer automatic matching unit 22 200527328 32 33 4 41 Manual comparison operation unit ticket issuance Unit blacklist management unit blacklist database
23twenty three