TWI777223B - Unmanned aerial vehicle traffic survey system and method thereof - Google Patents
Unmanned aerial vehicle traffic survey system and method thereof Download PDFInfo
- Publication number
- TWI777223B TWI777223B TW109128500A TW109128500A TWI777223B TW I777223 B TWI777223 B TW I777223B TW 109128500 A TW109128500 A TW 109128500A TW 109128500 A TW109128500 A TW 109128500A TW I777223 B TWI777223 B TW I777223B
- Authority
- TW
- Taiwan
- Prior art keywords
- image data
- data
- traffic
- converted
- uav
- Prior art date
Links
Images
Abstract
Description
本發明係有關於一種無人機交通調查系統,特別是一種結合人工智慧的無人機交通調查系統。本發明還涉及此無人機交通調查系統的無人機交通調查方法。The invention relates to an unmanned aerial vehicle traffic investigation system, in particular to an unmanned aerial vehicle traffic investigation system combined with artificial intelligence. The invention also relates to the UAV traffic investigation method of the UAV traffic investigation system.
交通調查(Traffic Survey)是指在道路現場蒐集調查地點之車流特性包括流率、速率、密度等資料。交通調查調查目的是要了解道路系統等交通特性,作為評估服務水準、容量分析、研擬交通改善計畫之用。Traffic survey refers to the collection of traffic characteristics including flow rate, speed, density and other data at the road site. The purpose of the traffic survey is to understand the traffic characteristics such as the road system, which is used to evaluate the service level, analyze the capacity, and develop the traffic improvement plan.
一般而言,高速公路已有佈設偵測器,其可長時間的蒐集交通數據。然而,這些偵測器的佈設位置固定,不一定能滿足研究地點的分析需求,故經常需要進行現場調查,進一步蒐集車流特性資料作分析之用。傳統的現場調查方式一般採用路側攝影機拍攝,需要施工將攝影機掛載在路側設施上,攝影機於調查結束後回收,再透過人工方式觀察影片進行資料登錄作業。現場調查之缺點有三點:(1)攝影機的佈設位置受限於現場既有之路側設施,不一定能完全符合需求;(2)攝影機的佈設需要於現場進行施工,影響車流運作;(3)資料登錄耗費人力而且容易因觀察人員的視覺疲勞而產生調查誤差。Generally speaking, highways have already deployed detectors, which can collect traffic data for a long time. However, the location of these detectors is fixed, which may not meet the analysis needs of the research site. Therefore, it is often necessary to conduct on-site surveys to further collect traffic flow characteristic data for analysis. Traditional on-site survey methods generally use roadside cameras to shoot, requiring construction to mount cameras on roadside facilities. The cameras are recovered after the survey is completed, and then the videos are manually observed for data registration. There are three shortcomings in the on-site investigation: (1) the location of the cameras is limited by the existing roadside facilities on the site, which may not fully meet the requirements; (2) the deployment of the cameras needs to be constructed on site, which affects the operation of traffic flow; (3) Data registration is labor-intensive and prone to survey errors due to observers' visual fatigue.
無人機(UAV)能於高空定點拍攝大範圍的車流影像,解決路側攝影機的佈設問題,而所拍攝之正射影片,能運用人工智慧電腦視覺的演算法進行自動化分析,排除資料登錄人為因素造成的誤差。以4K畫質錄影為例,拍攝範圍在700公尺以內尚能以人工或電腦視覺的方式有效的判別車輛。交通調查應涵蓋尖峰時段2-4小時,然而目前之市售無人機或長效無人機,其續航力約由30分鐘至80分鐘不等,單架次的飛行拍攝尚無法滿足調查需求。如採用多架次無人機進行連續拍攝,則會因每架次的拍攝可能因無人機懸停高度、拍攝角度設定有稍微差異,或因機型或攝影機規格不相同等因素,所取得之影片成像也不完全相同,無法將資料有效地整合以進行車流特性分析。Unmanned aerial vehicle (UAV) can shoot a large-scale traffic flow image at high altitude fixed point, solve the problem of roadside camera layout, and the orthophoto film shot can be automatically analyzed by using artificial intelligence computer vision algorithm to eliminate the human factors caused by data logging. error. Taking 4K video as an example, the vehicle can still be effectively identified by manual or computer vision within a shooting range of 700 meters. The traffic survey should cover 2-4 hours during peak hours. However, the current commercial drones or long-term drones have an endurance ranging from about 30 minutes to 80 minutes, and a single flight shooting cannot meet the survey needs. If multiple drones are used for continuous shooting, the video images obtained may be slightly different due to the hovering height of the drone, the setting of the shooting angle, or the different models or camera specifications. They are not identical, and the data cannot be effectively integrated for the analysis of traffic flow characteristics.
根據本發明之一實施例,本發明提出一種無人機交通調查系統,其包含第一無人機、第二無人機及計算模組。第一無人機於第一時間區間內拍攝一道路區段以獲得第一影像資料。第二無人機於第二時間區間內拍攝此道路區段以獲得第二影像資料。計算模組執行投影變換方法將第一影像資料及第二影像資料轉換為基於真實座標系統之第一轉換影像資料及第二轉換影像資料,再根據第一轉換影像資料及第二轉換影像資料計算此道路區段的車流特性資料。According to an embodiment of the present invention, the present invention provides a UAV traffic investigation system, which includes a first UAV, a second UAV, and a computing module. The first drone captures a road section within the first time interval to obtain first image data. The second drone captures the road section within the second time interval to obtain second image data. The computing module executes a projection transformation method to convert the first image data and the second image data into the first converted image data and the second converted image data based on the real coordinate system, and then calculates according to the first converted image data and the second converted image data Traffic characteristics data for this road segment.
在一實施例中,第一時間區間與第二時間區間為連續或部份重疊。In one embodiment, the first time interval and the second time interval are continuous or partially overlapping.
在一實施例中,計算模組執行投影變換方法將第一影像資料之影像座標系統及第二影像資料之影像座標系統轉換為真實座標系統以產生第一轉換影像資料及第二轉換影像資料。In one embodiment, the computing module performs a projective transformation method to convert the image coordinate system of the first image data and the image coordinate system of the second image data into a real coordinate system to generate the first transformed image data and the second transformed image data.
在一實施例中,計算模組分別由第一影像資料及第二影像資料之包含此道路區段之軌跡影像底圖中選擇複數個興趣點,並記錄該些興趣點於第一影像資料及第二影像資料中的影像座標及對應的經緯度座標,再執行投影變換方法將第一影像資料之影像座標系統及第二影像資料之影像座標系統轉換為真實座標系統。In one embodiment, the computing module selects a plurality of points of interest from the first image data and the second image data respectively from the track image base map including the road section, and records the points of interest in the first image data and The image coordinates and the corresponding latitude and longitude coordinates in the second image data, and then perform a projection transformation method to convert the image coordinate system of the first image data and the image coordinate system of the second image data into a real coordinate system.
在一實施例中,計算模組計算第一轉換影像資料及第二轉換影像資料之包含此道路區段之軌跡影像底圖的複數個車道方程式,再根據該些車道方程式計算此道路區段的車流特性資料。In one embodiment, the calculation module calculates a plurality of lane equations of the first converted image data and the second converted image data including the base map of the track image of the road section, and then calculates the road section according to the lane equations. Traffic characteristics data.
在一實施例中,計算模組分別由第一轉換影像資料及第二轉換影像資料之包含此道路區段之軌跡影像底圖的複數個車道線選擇複數個影像座標點,並執行多項式曲線擬合校估以獲得各個車道線的車道方程式。In one embodiment, the calculation module selects a plurality of image coordinate points from a plurality of lane lines of the first converted image data and the second converted image data including the track image base map of the road section, and performs polynomial curve simulation. Combined evaluation to obtain lane equations for each lane line.
在一實施例中,車流特性資料包含流率資料、速率資料、密度資料及車道變換資料中之一或以上。In one embodiment, the traffic characteristic data includes one or more of flow rate data, velocity data, density data, and lane change data.
根據本發明之另一實施例,本發明提出一種無人機交通調查方法,其包含下列步驟:由第一無人機於第一時間區間內拍攝一道路區段以獲得第一影像資料;透過第二無人機於第二時間區間內拍攝此道路區段以獲得第二影像資料;經由計算模組執行投影變換方法將第一影像資料及第二影像資料轉換為基於真實座標系統之第一轉換影像資料及第二轉換影像資料;以及透過計算模組根據第一轉換影像資料及第二轉換影像資料計算此道路區段的車流特性資料。According to another embodiment of the present invention, the present invention provides a UAV traffic investigation method, which includes the following steps: photographing a road section within a first time interval by a first UAV to obtain first image data; The drone captures the road section in the second time interval to obtain the second image data; the projection transformation method is performed by the computing module to convert the first image data and the second image data into the first transformed image data based on the real coordinate system and the second converted image data; and calculating the traffic flow characteristic data of the road section according to the first converted image data and the second converted image data through the computing module.
在一實施例中,第一時間區間與第二時間區間為連續或部份重疊。In one embodiment, the first time interval and the second time interval are continuous or partially overlapping.
在一實施例中,經由計算模組執行投影變換方法將第一影像資料及第二影像資料轉換為基於真實座標系統之第一轉換影像資料及第二轉換影像資料之步驟更包含下列步驟:經由計算模組執行投影變換方法將第一影像資料之影像座標系統及第二影像資料之影像座標系統轉換為真實座標系統以產生第一轉換影像資料及第二轉換影像資料。In one embodiment, the step of converting the first image data and the second image data into the first converted image data and the second converted image data based on the real coordinate system by executing the projective transformation method by the computing module further includes the following steps: The computing module executes a projective transformation method to convert the image coordinate system of the first image data and the image coordinate system of the second image data into a real coordinate system to generate the first transformed image data and the second transformed image data.
在一實施例中,經由計算模組執行投影變換方法將第一影像資料之影像座標系統及第二影像資料之影像座標系統轉換為真實座標系統以產生第一轉換影像資料及第二轉換影像資料之步驟更包含下列步驟:透過計算模組分別由第一影像資料及第二影像資料之包含此道路區段之軌跡影像底圖中選擇複數個興趣點,並記錄該些興趣點於第一影像資料及第二影像資料中的影像座標及對應的經緯度座標;以及經由計算模組分執行投影變換方法將第一影像資料之影像座標系統及第二影像資料之影像座標系統轉換為真實座標系統。In one embodiment, the projection transformation method is performed by the computing module to transform the image coordinate system of the first image data and the image coordinate system of the second image data into a real coordinate system to generate the first transformed image data and the second transformed image data The step further includes the following steps: selecting a plurality of points of interest from the first image data and the second image data in the track image base map including the road section respectively, and recording these points of interest in the first image The image coordinates and the corresponding latitude and longitude coordinates in the data and the second image data; and the projection transformation method is performed by the computing module to convert the image coordinate system of the first image data and the image coordinate system of the second image data into a real coordinate system.
在一實施例中,透過計算模組根據第一轉換影像資料及第二轉換影像資料計算此道路區段的車流特性資料之步驟更包含下列步驟:由計算模組計算第一轉換影像資料及第二轉換影像資料之包含此道路區段之軌跡影像底圖的複數個車道方程式,再根據該些車道方程式計算此道路區段的車流特性資料。In one embodiment, the step of calculating the traffic flow characteristic data of the road section according to the first converted image data and the second converted image data by the calculation module further includes the following steps: calculating, by the calculation module, the first converted image data and the second converted image data; 2. Convert a plurality of lane equations of the image data including the base map of the track image of the road section, and then calculate the traffic flow characteristic data of the road section according to the lane equations.
在一實施例中,由計算模組計算第一轉換影像資料及第二轉換影像資料之包含此道路區段之軌跡影像底圖的該些車道方程式之步驟更包含下列步驟:透過計算模組分別由第一轉換影像資料及第二轉換影像資料之包含此道路區段之軌跡影像底圖的複數個車道線選擇複數個影像座標點,並執行多項式曲線擬合校估以獲得各個車道線的該車道方程式。In one embodiment, the step of calculating, by the calculation module, the lane equations of the first converted image data and the second converted image data including the base map of the track image of the road section further includes the following steps: respectively, through the calculation module Select a plurality of image coordinate points from a plurality of lane lines of the first converted image data and the second converted image data including the track image base map of the road section, and perform polynomial curve fitting and calibration to obtain the corresponding lane lines of each lane line. Lane Equation.
在一實施例中,車流特性資料包含流率資料、速率資料、密度資料及車道變換資料中之一或以上。In one embodiment, the traffic characteristic data includes one or more of flow rate data, velocity data, density data, and lane change data.
承上所述,依本發明之無人機交通調查系統及其交通調查方法,其可具有一或多個下述優點:Based on the above, according to the UAV traffic survey system and the traffic survey method thereof of the present invention, it can have one or more of the following advantages:
(1)本發明之一實施例中,無人機交通調查系統採用無人機進行交通調查且整合人工智慧技術,故可完全符合交通調查的需求,且不會影響車流運作,並可大幅降低人力的需求,有效地改善了現場調查的缺點。(1) In one embodiment of the present invention, the UAV traffic investigation system adopts UAV to conduct traffic investigation and integrates artificial intelligence technology, so it can fully meet the needs of traffic investigation, and will not affect the operation of traffic flow, and can greatly reduce the labor cost Demand, effectively improving the shortcomings of on-site investigations.
(2)本發明之一實施例中,無人機交通調查系統採用多架無人機進行輪替拍攝調查,可提供大範圍與長時間的拍攝需求,且可確保拍攝之交通影像能保持連續不間斷,故能確實滿足交通調查的需求。(2) In one embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to conduct alternate shooting investigations, which can provide a wide range and long-term shooting requirements, and can ensure that the captured traffic images can be kept continuous and uninterrupted. , so it can indeed meet the needs of traffic surveys.
(3)本發明之一實施例中,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且整合攝影測量方法進行影像座標與真實座標轉換,使該些無人機取得之交通調查資料能具有相同的座標系統以達到時空同步,有效地提升交通調查的準確性及資料整理的作業效率。(3) In one embodiment of the present invention, the UAV traffic survey system uses multiple UAVs to conduct alternate shooting surveys, and integrates photogrammetry methods to convert image coordinates and real coordinates, so that the traffic survey data obtained by these UAVs It can have the same coordinate system to achieve time-space synchronization, effectively improving the accuracy of traffic surveys and the efficiency of data sorting.
(4)本發明之一實施例中,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且運用人工智慧電腦視覺的演算法進行自動化分析,故能有效地排除資料登錄的人為誤差,進一步提升交通調查的準確性。(4) In one embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to conduct alternate shooting investigations, and uses artificial intelligence computer vision algorithms to perform automated analysis, so it can effectively eliminate human errors in data registration , to further improve the accuracy of traffic surveys.
(5)本發明之一實施例中,無人機交通調查系統設計巧妙,且有效地整合多項技術,故能在不大幅增加成本的前提下達到所欲達到的功效,故具實用性且應用上更為廣泛。(5) In one embodiment of the present invention, the UAV traffic investigation system is ingeniously designed and effectively integrates multiple technologies, so it can achieve the desired effect without greatly increasing the cost, so it is practical and applicable. more extensive.
以下將參照相關圖式,說明依本發明之無人機交通調查系統及其交通調查方法之實施例,為了清楚與方便圖式說明之故,圖式中的各部件在尺寸與比例上可能會被誇大或縮小地呈現。為使便於理解,下述實施例中之相同元件係以相同之符號標示來說明。Embodiments of the UAV traffic survey system and the traffic survey method according to the present invention will be described below with reference to the relevant drawings. For the sake of clarity and convenience in the description of the drawings, the dimensions and proportions of the components in the drawings may be changed. Presented exaggerated or reduced. For ease of understanding, the same elements in the following embodiments are denoted by the same symbols.
請參閱第1圖,其係為本發明之一實施例之無人機交通調查系統之第一示意圖。如圖所示,無人機交通調查系統1包含第一無人機11、第二無人機12及計算模組13。在另一實施例中,無人機交通調查系統1也可包含三個以上的無人機。Please refer to FIG. 1 , which is a first schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention. As shown in the figure, the UAV
第一無人機11於第一時間區間T1內拍攝道路區段以獲得第一影像資料,並傳送至計算模組13。當第一無人機11的電力低於預設電力門檻值時,第二無人機12開始升空以接替第一無人機11;當第二無人機12升空至相同高度接替第一無人機11時,第一無人機11降落至地面,並更換電池待命。其中,計算模組13可為各種電腦裝置,如伺服器、工作站及筆記型電腦等。The
第二無人機12於第二時間區間T2內拍攝道路區段以獲得第二影像資料,並傳送至計算模組13。同樣的,當第二無人機12的電力低於預設電力門檻值時,第一無人機11開始升空以接替第二無人機12;當第一無人機11升空至相同高度接替第二無人機12時,第二無人機12降落至地面,並更換電池待命,以準備在一段時間再次接替第一無人機11。The
在後續的第三時間區間T3、第四時間區間T4及之後的時間區間均重覆上述的機制,且其相鄰的時間區間為連續或部份重疊;例如,第二時間區間T2與第一時間區間T1為連續或部份重疊;第三時間區間T3與第二時間區間T2為連續或部份重疊;第四時間區間T4與第三時間區間T3為連續或部份重疊。預設電力門檻值需足夠使第一無人機11及第二無人機12等待接替並安全降落。透過上述的機制,無人機交通調查系統1可提供大範圍與長時間的拍攝需求,且可確保拍攝之交通影像能保持連續不間斷。The above-mentioned mechanism is repeated in the subsequent third time interval T3, fourth time interval T4 and subsequent time intervals, and the adjacent time intervals are continuous or partially overlapping; for example, the second time interval T2 and the first time interval T2 The time interval T1 is continuous or partially overlapping; the third time interval T3 and the second time interval T2 are continuous or partially overlapping; the fourth time interval T4 and the third time interval T3 are continuous or partially overlapping. The preset power threshold should be sufficient for the
請參閱第2圖,其係為本發明之一實施例之無人機交通調查系統之第二示意圖。如圖所示,道路區段R包含加速車道L0、第一車道L1、第二車道L2及第三車道L3。計算模組13由第一影像資料包含道路區段R的軌跡影像底圖中選擇複數個興趣點POI,並記錄該些興趣點POI於第一影像資料中的影像座標及對應該些興趣點POI的影像座標的經緯度座標,再執行投影變換方法(Projective transformation between planes)將第一影像資料之影像座標系統為真實座標系統,以取得第一轉換影像資料,如式(1)及式(2)所示:Please refer to FIG. 2 , which is a second schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention. As shown, the road section R includes an acceleration lane L0, a first lane L1, a second lane L2, and a third lane L3. The
……………(1) ……………(1)
……………(2) ……………(2)
其中,(x, y)為影像座標;(X, Y)為真實座標;e0、e1、e2、f0、f1、f2、g1、g2為轉換參數,該些轉換參數可由校估而得。Among them, (x, y) are image coordinates; (X, Y) are real coordinates; e0, e1, e2, f0, f1, f2, g1, g2 are conversion parameters, which can be obtained by calibration.
同樣的,計算模組13也透過相同的方式取得第二轉換影像資料。透過此機制,計算模組13能將第一影像資料及第二影像資料轉換為基於同一個真實座標系統之第一轉換影像資料及第二轉換影像資料,使第一無人機11及第二無人機12拍攝的影像能夠在空間同步。由於第一無人機11及第二無人機12為輪替拍攝,故可能因第一無人機11及第二無人機12的懸停高度、拍攝角度設定或因其它因素,導致二者取得之影片成像可能有些微差異;然而,上述的機制可以有效地修正此差異,使第一無人機11及第二無人機12拍攝的影像能夠在空間上確實達到同步。Similarly, the
請參閱第3圖,其係為本發明之一實施例之無人機交通調查系統之第三示意圖。如圖所示,接下來,計算模組13由第一轉換影像資料包含此道路區段之軌跡影像底圖的複數個車道線中的每一個車道線選擇複數個影像座標點,並執行多項式曲線擬合(Curve fitting)校估以獲得該些車道線的車道方程式LF1~LF6;在本實施例中,計算模組13由第一轉換影像資料包含此道路區段之軌跡影像底圖的複數個車道線中的每一個車道線選擇10個以上的影像座標點,並執行多項式曲線擬合校估以獲得該些車道線的車道方程式LF1~LF6。其中,車道方程式LF1及車道方程式LF2之間的車道為加速車道L0;車道方程式LF3及車道方程式LF4之間的車道為第一車道L1;車道方程式LF4及車道方程式LF5之間的車道為第二車道L2;車道方程式LF5及車道方程式LF6之間的車道為第三車道L3。取得該些車道線的車道方程式LF1~LF6後,計算模組13則可根據各個車輛的車輛軌跡的車頭中心之影像座標判斷車輛所在的車道位置;如此即可獲得此道路區段在第一轉換影像資料之時間區間的車流特性資料;其中,車流特性資料可包含流率資料、速率資料、密度資料及車道變換資料中之一或以上。Please refer to FIG. 3 , which is a third schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention. As shown in the figure, next, the
同樣的,計算模組13也可透過相同的方式由第二轉換影像資料包含此道路區段之軌跡影像底圖的複數個車道線中的每一個車道線選擇複數個影像座標點,並執行多項式曲線擬合校估以獲得該些車道線的車道方程式LF1~LF6,再根據各個車輛的車輛軌跡的車頭中心之影像座標判斷車輛所在的車道位置,即可以獲得此道路區段在第二轉換影像資料之時間區間的車流特性資料。由上述可知,本實施例之無人機交通調查系統1可透過人工智慧(AI)影像分析來進行道路區段的交通調查,透過第一無人機11及第二無人機12於道路區段上方拍攝影像,再運用深度學習演算法分析影像中之車流進行車輛的偵測、追蹤、分類,藉此取得車流特性資料,且可達到時空同步的效果。Similarly, the
請參閱第4A圖~第4D圖,其係為本發明之一實施例之無人機交通調查系統之量測道路區段之半小時車流軌跡時空圖。第4A圖~第4D圖為透過本實施例的機制取得的一道路區段的各個車道的半小時(07:00~07:30)車流軌跡時空圖;第4A圖表示第三車道L3的半小時車流軌跡時空圖;第4B圖表示第二車道L2的半小時車流軌跡時空圖;第4C圖表示第一車道L1的半小時車流軌跡時空圖;第4D圖表示加速車道的半小時車流軌跡時空圖。各車道對應之位置如圖片左側簡圖所示;圖中軌跡線代表每台車輛,軌跡線的顏色則代表該時間點下車輛的速率,而數值則參照圖片右側顏色條。Please refer to FIGS. 4A to 4D , which are spatiotemporal diagrams of half-hour traffic trajectories of road sections measured by the UAV traffic survey system according to an embodiment of the present invention. Figures 4A to 4D are the half-hour (07:00~07:30) traffic trajectories of each lane of a road section obtained through the mechanism of this embodiment; Figure 4A shows the half-hour (07:00~07:30) traffic trajectories of the third lane L3. Hourly traffic trajectory spatio-temporal map; Figure 4B represents the half-hour traffic trajectory spatiotemporal map of the second lane L2; Figure 4C represents the half-hour traffic trajectory spatiotemporal map of the first lane L1; Figure 4D represents the half-hour traffic trajectory spatio-temporal map of the acceleration lane picture. The corresponding position of each lane is shown in the diagram on the left side of the picture; the trajectory line in the picture represents each vehicle, the color of the trajectory line represents the speed of the vehicle at that time point, and the value refers to the color bar on the right side of the picture.
請參閱第5A圖~第5C圖,其係為本發明之一實施例之無人機交通調查系統之量測道路區段之流率-時間圖、密度-時間圖及速率-時間圖。Please refer to FIGS. 5A to 5C, which are flow rate-time diagrams, density-time diagrams, and rate-time diagrams of the measured road section of the UAV traffic survey system according to an embodiment of the present invention.
如第5A圖所示,曲線F0表示加速車道之流率-時間的變化;曲線F1表示第一車道L1之流率-時間的變化;曲線F2表示第二車道L2之流率-時間的變化;曲線F3表示第三車道L3之流率-時間的變化;曲線FT表示總和。As shown in Figure 5A, the curve F0 represents the flow rate-time change of the acceleration lane; the curve F1 represents the flow rate-time change of the first lane L1; the curve F2 represents the flow rate-time change of the second lane L2; Curve F3 represents the flow rate-time variation of the third lane L3; curve FT represents the sum.
如第5B圖所示,曲線D0表示加速車道之密度-時間的變化;曲線D1表示第一車道L1之密度-時間的變化;曲線D2表示第二車道L2之密度-時間的變化;曲線D3表示第三車道L3之密度-時間的變化;曲線DA表示主線平均。As shown in Fig. 5B, the curve D0 represents the density-time variation of the acceleration lane; the curve D1 represents the density-time variation of the first lane L1; the curve D2 represents the density-time variation of the second lane L2; the curve D3 represents Density-time variation of the third lane L3; curve DA represents the main line average.
如第5C圖所示,曲線S0表示加速車道之速率-時間的變化;曲線S1表示第一車道L1之速率-時間的變化;曲線S2表示第二車道L2之速率-時間的變化;曲線S3表示第三車道L3之速率-時間的變化;曲線SA表示主線平均。As shown in Figure 5C, the curve S0 represents the speed-time change of the acceleration lane; the curve S1 represents the speed-time change of the first lane L1; the curve S2 represents the speed-time change of the second lane L2; the curve S3 represents Velocity-time variation of the third lane L3; curve SA represents the main line average.
由上述可知,本實施例之無人機交通調查系統1確實可達到時空同步的功效,確實滿足交通調查的需求,且有效地提升交通調查的準確性及資料整理的作業效率。It can be seen from the above that the UAV
值得一提的是,現有的現場調查方式由於各種技術上的限制,使其無法完全符合交通調查的需求,且容易影響車流運作,並容易產生人為的調查誤差。相反的,根據本發明之實施例,無人機交通調查系統採用無人機進行交通調查且整合人工智慧技術,故可完全符合交通調查的需求,且不會影響車流運作,並可大幅降低人力的需求,有效地改善了現場調查的缺點。It is worth mentioning that due to various technical limitations, the existing on-site survey methods cannot fully meet the needs of traffic surveys, and easily affect the operation of traffic flow, and are prone to human survey errors. On the contrary, according to the embodiment of the present invention, the UAV traffic investigation system adopts UAV to conduct traffic investigation and integrates artificial intelligence technology, so it can fully meet the needs of traffic investigation without affecting the operation of traffic flow, and can greatly reduce the demand for manpower , effectively improving the shortcomings of on-site investigation.
現有的無人機交通調查方法由於受限於無人機的續航力,故無法滿足交通調查的需求。相反的,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,可提供大範圍與長時間的拍攝需求,且可確保拍攝之交通影像能保持連續不間斷,故能確實滿足交通調查的需求。The existing UAV traffic survey methods cannot meet the needs of traffic survey due to the limitation of the endurance of the UAV. On the contrary, according to the embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to perform alternate shooting investigations, which can provide a wide range and long-term shooting requirements, and can ensure that the captured traffic images can be kept continuous and uninterrupted. Therefore, it can indeed meet the needs of traffic surveys.
現有的無人機交通調查方法可能因懸停高度及拍攝角度設定之差異,或因機型或攝影機規格不相同等因素,導致無法將資料有效地整合以進行車流特性分析。相反的,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且整合攝影測量方法進行影像座標與真實座標轉換,使該些無人機取得之交通調查資料能具有相同的座標系統以達到時空同步,有效地提升交通調查的準確性及資料整理的作業效率。The existing UAV traffic investigation methods may not be able to effectively integrate the data for analysis of traffic flow characteristics due to differences in hovering height and shooting angle settings, or due to different models or camera specifications. On the contrary, according to the embodiment of the present invention, the UAV traffic survey system uses multiple UAVs to perform alternate shooting surveys, and integrates the photogrammetry method to convert image coordinates and real coordinates, so that the traffic survey data obtained by these UAVs can be It has the same coordinate system to achieve time-space synchronization, which effectively improves the accuracy of traffic surveys and the efficiency of data sorting.
另外,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且運用人工智慧電腦視覺的演算法進行自動化分析,故能有效地排除資料登錄的人為誤差,進一步提升交通調查的準確性。In addition, according to the embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to conduct alternate shooting investigations, and uses artificial intelligence computer vision algorithms to perform automatic analysis, so it can effectively eliminate human errors in data registration, and further Improve the accuracy of traffic surveys.
再者,根據本發明之實施例,無人機交通調查系統設計巧妙,且有效地整合多項技術,故能在不大幅增加成本的前提下達到所欲達到的功效,故具實用性且應用上更為廣泛。由上述可知,本發明之實施例之無人機交通調查系統及其方法確實能達到無法預期之功效。Furthermore, according to the embodiment of the present invention, the UAV traffic investigation system is ingeniously designed and effectively integrates multiple technologies, so that the desired effect can be achieved without greatly increasing the cost, so it is practical and more applicable. for broad. As can be seen from the above, the UAV traffic investigation system and method according to the embodiments of the present invention can indeed achieve unexpected effects.
請參閱第6圖,其係為本發明之一實施例之無人機交通調查方法之流程圖。本實施例之無人機交通調查方法包含下列步驟:Please refer to FIG. 6 , which is a flowchart of a method for investigating drone traffic according to an embodiment of the present invention. The UAV traffic investigation method of this embodiment includes the following steps:
步驟S61:由第一無人機於第一時間區間內拍攝一道路區段以獲得第一影像資料。Step S61 : photographing a road section within the first time interval by the first drone to obtain the first image data.
步驟S62:透過第二無人機於第二時間區間內拍攝此道路區段以獲得第二影像資料。Step S62 : photographing the road section within the second time interval by the second drone to obtain the second image data.
步驟S63:經由計算模組執行投影變換方法將第一影像資料及第二影像資料轉換為基於真實座標系統之第一轉換影像資料及第二轉換影像資料。Step S63 : Convert the first image data and the second image data into the first converted image data and the second converted image data based on the real coordinate system by performing a projective transformation method through the computing module.
步驟S64:透過計算模組根據第一轉換影像資料及第二轉換影像資料計算此道路區段的車流特性資料。Step S64: Calculate the traffic flow characteristic data of the road section according to the first converted image data and the second converted image data through the calculation module.
綜上所述,根據本發明之實施例,無人機交通調查系統採用無人機進行交通調查且整合人工智慧技術,故可完全符合交通調查的需求,且不會影響車流運作,並可大幅降低人力的需求,有效地改善了現場調查的缺點。To sum up, according to the embodiments of the present invention, the UAV traffic investigation system uses UAVs for traffic investigation and integrates artificial intelligence technology, so it can fully meet the needs of traffic investigation without affecting the operation of traffic flow, and can greatly reduce manpower demand, effectively improving the shortcomings of on-site investigations.
又,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,可提供大範圍與長時間的拍攝需求,且可確保拍攝之交通影像能保持連續不間斷,故能確實滿足交通調查的需求。In addition, according to the embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to perform alternate shooting surveys, which can provide a wide range and long-term shooting requirements, and can ensure that the captured traffic images can be kept continuous and uninterrupted, so It can really meet the needs of traffic surveys.
此外,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且整合攝影測量方法進行影像座標與真實座標轉換,使該些無人機取得之交通調查資料能具有相同的座標系統以達到時空同步,有效地提升交通調查的準確性及資料整理的作業效率。In addition, according to an embodiment of the present invention, the UAV traffic survey system uses multiple UAVs to perform alternate shooting surveys, and integrates the photogrammetry method to convert image coordinates and real coordinates, so that the traffic survey data obtained by these UAVs can have The same coordinate system can achieve the synchronization of time and space, which can effectively improve the accuracy of traffic survey and the efficiency of data processing.
另外,根據本發明之實施例,無人機交通調查系統採用多架無人機進行輪替拍攝調查,且運用人工智慧電腦視覺的演算法進行自動化分析,故能有效地排除資料登錄的人為誤差,進一步提升交通調查的準確性。In addition, according to the embodiment of the present invention, the UAV traffic investigation system uses multiple UAVs to conduct alternate shooting investigations, and uses artificial intelligence computer vision algorithms to perform automatic analysis, so it can effectively eliminate human errors in data registration, and further Improve the accuracy of traffic surveys.
再者,根據本發明之實施例,無人機交通調查系統設計巧妙,且有效地整合多項技術,故能在不大幅增加成本的前提下達到所欲達到的功效,故具實用性且應用上更為廣泛。Furthermore, according to the embodiment of the present invention, the UAV traffic investigation system is ingeniously designed and effectively integrates multiple technologies, so that the desired effect can be achieved without greatly increasing the cost, so it is practical and more applicable. for broad.
可見本發明在突破先前之技術下,確實已達到所欲增進之功效,且也非熟悉該項技藝者所易於思及,其所具之進步性、實用性,顯已符合專利之申請要件,爰依法提出專利申請,懇請 貴局核准本件發明專利申請案,以勵創作,至感德便。It can be seen that the present invention has indeed achieved the desired enhancement effect under the breakthrough of the previous technology, and it is not easy for those who are familiar with the technology to think about it. Yuan has filed a patent application in accordance with the law, and I implore your bureau to approve this invention patent application, so as to encourage creation, and to be grateful.
以上所述僅為舉例性,而非為限制性者。其它任何未脫離本發明之精神與範疇,而對其進行之等效修改或變更,均應該包含於後附之申請專利範圍中。The above description is exemplary only, not limiting. Any other equivalent modifications or changes without departing from the spirit and scope of the present invention should be included in the appended patent application scope.
1:無人機交通調查系統 11:第一無人機 12:第二無人機 13:計算模組 T1:第一時間區間 T2:第二時間區間 T3:第三時間區間 T4:第四時間區間 R:道路區段 L0:加速車道 L1:第一車道 L2:第二車道 L3:第三車道 POI:興趣點 LF1~LF6:車道方程式 F0~F3,FT,D0~D3,DA,S0~S3,SA:曲線 S61~S64:步驟流程1: UAV Traffic Survey System 11: The first drone 12: Second Drone 13: Computing Module T1: The first time interval T2: Second time interval T3: The third time interval T4: Fourth time interval R: road segment L0: acceleration lane L1: first lane L2: Second lane L3: Third lane POI: Point of Interest LF1~LF6: Lane Equation F0~F3,FT,D0~D3,DA,S0~S3,SA: Curve S61~S64: Step flow
第1圖 係為本發明之一實施例之無人機交通調查系統之第一示意圖。FIG. 1 is a first schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention.
第2圖 係為本發明之一實施例之無人機交通調查系統之第二示意圖。FIG. 2 is a second schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention.
第3圖 係為本發明之一實施例之無人機交通調查系統之第三示意圖。FIG. 3 is a third schematic diagram of a UAV traffic investigation system according to an embodiment of the present invention.
第4A圖~第4D圖 係為本發明之一實施例之無人機交通調查系統之量測道路區段之半小時車流軌跡時空圖。Figures 4A to 4D are spatiotemporal diagrams of half-hour traffic trajectories of road sections measured by the UAV traffic survey system according to an embodiment of the present invention.
第5A圖 係為本發明之一實施例之無人機交通調查系統之量測道路區段之流率-時間圖。FIG. 5A is a flow rate-time diagram of the measured road section of the UAV traffic survey system according to an embodiment of the present invention.
第5B圖 係為本發明之一實施例之無人機交通調查系統之量測道路區段之密度-時間圖。FIG. 5B is a density-time diagram of the measured road section of the UAV traffic survey system according to an embodiment of the present invention.
第5C圖 係為本發明之一實施例之無人機交通調查系統之量測道路區段之速率-時間圖。FIG. 5C is a velocity-time diagram of the measured road section of the UAV traffic survey system according to an embodiment of the present invention.
第6圖 係為本發明之一實施例之無人機交通調查方法之流程圖。FIG. 6 is a flow chart of a method for investigating drone traffic according to an embodiment of the present invention.
1:無人機交通調查系統1: UAV Traffic Survey System
11:第一無人機11: The first drone
12:第二無人機12: Second Drone
13:計算模組13: Computing Module
T1:第一時間區間T1: The first time interval
T2:第二時間區間T2: Second time interval
T3:第三時間區間T3: The third time interval
T4:第四時間區間T4: Fourth time interval
Claims (14)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109128500A TWI777223B (en) | 2020-08-21 | 2020-08-21 | Unmanned aerial vehicle traffic survey system and method thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109128500A TWI777223B (en) | 2020-08-21 | 2020-08-21 | Unmanned aerial vehicle traffic survey system and method thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
TW202209270A TW202209270A (en) | 2022-03-01 |
TWI777223B true TWI777223B (en) | 2022-09-11 |
Family
ID=81746928
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW109128500A TWI777223B (en) | 2020-08-21 | 2020-08-21 | Unmanned aerial vehicle traffic survey system and method thereof |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI777223B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI438728B (en) * | 2012-04-25 | 2014-05-21 | Hon Hai Prec Ind Co Ltd | System and method for controlling traffic flow information |
WO2017030737A1 (en) * | 2015-08-20 | 2017-02-23 | Motionloft, Inc. | Object detection and analysis via unmanned aerial vehicle |
CN106781579A (en) * | 2017-02-24 | 2017-05-31 | 南京林业大学 | Interim unmanned plane traffic lights and intersection monitoring system and control method |
TW201822067A (en) * | 2016-12-08 | 2018-06-16 | 財團法人車輛研究測試中心 | System and method for detecting lane lines of road enhancing the recognition of fuzzy lane lines, a single lane line and far lane lines |
CN111145545A (en) * | 2019-12-25 | 2020-05-12 | 西安交通大学 | Road traffic behavior unmanned aerial vehicle monitoring system and method based on deep learning |
-
2020
- 2020-08-21 TW TW109128500A patent/TWI777223B/en active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI438728B (en) * | 2012-04-25 | 2014-05-21 | Hon Hai Prec Ind Co Ltd | System and method for controlling traffic flow information |
WO2017030737A1 (en) * | 2015-08-20 | 2017-02-23 | Motionloft, Inc. | Object detection and analysis via unmanned aerial vehicle |
TW201822067A (en) * | 2016-12-08 | 2018-06-16 | 財團法人車輛研究測試中心 | System and method for detecting lane lines of road enhancing the recognition of fuzzy lane lines, a single lane line and far lane lines |
CN106781579A (en) * | 2017-02-24 | 2017-05-31 | 南京林业大学 | Interim unmanned plane traffic lights and intersection monitoring system and control method |
CN111145545A (en) * | 2019-12-25 | 2020-05-12 | 西安交通大学 | Road traffic behavior unmanned aerial vehicle monitoring system and method based on deep learning |
Also Published As
Publication number | Publication date |
---|---|
TW202209270A (en) | 2022-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111145545B (en) | Road traffic behavior unmanned aerial vehicle monitoring system and method based on deep learning | |
US9798928B2 (en) | System for collecting and processing aerial imagery with enhanced 3D and NIR imaging capability | |
CN105898216B (en) | A kind of number method of counting carried out using unmanned plane | |
CN107194989B (en) | Traffic accident scene three-dimensional reconstruction system and method based on unmanned aerial vehicle aircraft aerial photography | |
WO2020062434A1 (en) | Static calibration method for external parameters of camera | |
Barmpounakis et al. | How accurate are small drones for measuring microscopic traffic parameters? | |
CN107504957A (en) | The method that three-dimensional terrain model structure is quickly carried out using unmanned plane multi-visual angle filming | |
CN106767706A (en) | A kind of unmanned plane reconnoitres the Aerial Images acquisition method and system of the scene of a traffic accident | |
Adams et al. | Unmanned aerial vehicle data acquisition for damage assessment in hurricane events | |
CN107545538B (en) | Panoramic image splicing method and device based on unmanned aerial vehicle | |
CN110189411A (en) | Emergency management and rescue Search Area method for visualizing after a kind of accident of aircraft | |
TWI444593B (en) | Ground target geolocation system and method | |
Cho et al. | Diml/cvl rgb-d dataset: 2m rgb-d images of natural indoor and outdoor scenes | |
CN110580443B (en) | Low-altitude near-real-time building earthquake damage assessment method | |
CN102930524A (en) | Method for detecting heads based on vertically-placed depth cameras | |
KR20160082886A (en) | Method and system for mapping using UAV and multi-sensor | |
CN113705349A (en) | Attention power analysis method and system based on sight estimation neural network | |
Alkaabi et al. | Applications of unmanned aerial vehicle (UAV) technology for research and education in UAE | |
Jacobs et al. | Cloud motion as a calibration cue | |
TWI777223B (en) | Unmanned aerial vehicle traffic survey system and method thereof | |
CN116309851B (en) | Position and orientation calibration method for intelligent park monitoring camera | |
Domozi et al. | Surveying private pools in suburban areas with neural network based on drone photos | |
CN103438867A (en) | Onboard multi-vision three-dimensional aviation remote sensing system of small-size light unmanned plane | |
CN114120236A (en) | Method for identifying and positioning low-altitude target | |
CN203443586U (en) | Small light UAV-mounted (unmanned aerial vehicle-mounted) multi-view three-dimensional aerial remote sensing system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GD4A | Issue of patent certificate for granted invention patent |