TWI782806B - Point cloud rendering method - Google Patents
Point cloud rendering method Download PDFInfo
- Publication number
- TWI782806B TWI782806B TW110144931A TW110144931A TWI782806B TW I782806 B TWI782806 B TW I782806B TW 110144931 A TW110144931 A TW 110144931A TW 110144931 A TW110144931 A TW 110144931A TW I782806 B TWI782806 B TW I782806B
- Authority
- TW
- Taiwan
- Prior art keywords
- data
- distance
- pixel
- sub
- frame
- Prior art date
Links
Images
Landscapes
- Image Generation (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
Description
本發明是關於一種點雲算圖方法，特別是關於一種通過子像素資料、終像素資料或種子像素的設計，提升輸出畫面品質的點雲算圖方法。 The present invention relates to a point cloud calculation method, in particular to a point cloud calculation method for improving the quality of an output image by designing sub-pixel data, final pixel data or seed pixels.
點雲(Point Cloud)是呈現空間座標中立體形狀、物件或場景的資料集，通過3D掃描、光達、雷射測距掃描、影像軟體推算等方式建構點雲的資料。點雲的檔案中可包含標頭區(hoader)及資料區(data)，標頭區包含資料型態、類別標籤等資訊，資料區則為空間中各個點資料的集合，點資料可包含位置座標(例如XYZ三軸座標值)、顏色資料(例如RGB資料)或者其他如反射強度、法向等資訊。 Point Cloud is a data set that presents three-dimensional shapes, objects, or scenes in spatial coordinates. The point cloud data is constructed by means of 3D scanning, lidar, laser ranging scanning, and image software calculations. A point cloud file can contain a header area (hoader) and a data area (data). The header area contains information such as data type, category label, etc., and the data area is a collection of data of each point in the space. The point data can include the location Coordinates (such as XYZ three-axis coordinates), color data (such as RGB data) or other information such as reflection intensity and normal direction.
通過點雲建立的立體模型，可以具有許多不同的應用領域，例如古蹟維護、建築工程、自動駕駛、數位城市等，都能通過點雲資料的運用，達到提升操作效率及準確性等效果。另外，在影視或動畫領域上，也可通過點雲算圖的方式，產生所需的影像、場景等畫面，對於藝術領域的創作也能產生實際的影響。然而，現有的點雲算圖方式，不論是將點雲資料轉換成三角形網格模型再將網格匯入傳統算圖流程，或者直接以點雲資料算圖，都具有其侷限性。例如直接以點雲資料算圖，產生的畫面上常有顆粒感、波紋等瑕疵，無法達到 擬真的效果，在影視及動畫領域上難以達到畫面需求的水準，因此，在點雲資料的應用上仍具有相當的缺陷。 The three-dimensional model established through the point cloud can have many different application fields, such as monument maintenance, construction engineering, automatic driving, digital city, etc., and can achieve the effect of improving operation efficiency and accuracy through the use of point cloud data. In addition, in the field of film and television or animation, point cloud computing can also be used to generate the required images, scenes and other pictures, which can also have a practical impact on the creation of the art field. However, the existing point cloud calculation method, whether it is converting point cloud data into a triangular mesh model and then importing the grid into the traditional drawing process, or directly using point cloud data to calculate graphics, has its limitations. For example, directly using point cloud data to calculate the image, the generated image often has graininess, ripples and other defects, which cannot be achieved. The realistic effect is difficult to meet the level of picture requirements in the field of film and television and animation. Therefore, there are still considerable defects in the application of point cloud data.
有鑑於此，雖然目前已有以點雲資料算圖的方法，但現有的方法仍有其侷限性，且難以確保輸出成果的品質。對此，本發明之發明人思索並設計一種點雲算圖方法，針對現有技術之缺失加以改善，進而增進產業上之實施利用。 In view of this, although there are existing methods for calculating images from point cloud data, the existing methods still have their limitations, and it is difficult to ensure the quality of the output results. In this regard, the inventor of the present invention conceived and designed a point cloud calculation method to improve the deficiencies of the existing technology, thereby enhancing the implementation and utilization in the industry.
有鑑於上述習知技術之問題，本發明之目的就是在提供一種點雲算圖方法，以解決習知之點雲算圖方法在輸出畫面上容易產生雜點及波紋等瑕疵之問題。 In view of the above-mentioned problems in the prior art, the purpose of the present invention is to provide a point cloud calculation method to solve the problem that the conventional point cloud calculation method is prone to produce defects such as noise points and ripples on the output screen.
根據本發明之一目的，提出一種點雲算圖方法，藉由處理器存取記憶體中的點雲資料，執行計算程序以更新影格原稿，影格原稿所呈現的投影平面對應於觀察中心，點雲算圖方法包含以下步驟：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含預定數量的子像素資料；存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標；將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定；將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，***原本的子像素資料當中並依據相隔距離由低至高的順序進行排列；以及藉由點雲資料完成影格原稿當中各像素的更新。 According to an object of the present invention, a point cloud calculation method is proposed. The processor accesses the point cloud data in the memory and executes the calculation program to update the original frame. The projection plane presented by the original frame corresponds to the observation center, and the point The cloud calculation method includes the following steps: setting the frame original, the frame original contains a pixel matrix composed of a plurality of pixels, and each pixel contains a predetermined number of sub-pixel data; accessing the position data and color data of the point cloud data, calculating the position data and The distance between the observation center and the frame coordinates of the position data projected on the frame original; the frame coordinates are converted into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the frame original; the color of the point cloud data The data, separation distance and weight value are converted into sub-pixel data of each pixel, inserted into the original sub-pixel data and arranged according to the order of separation distance from low to high; and the update of each pixel in the original frame is completed by point cloud data.
較佳地，影格座標與像素座標之間的距離越近，權重值越高。 Preferably, the closer the distance between the frame coordinates and the pixel coordinates, the higher the weight value.
較佳地，當進行影格原稿的更新使子像素資料的數量超過預定數量時，可移除相隔距離最遠的子像素資料。 Preferably, when the update of the frame manuscript causes the number of sub-pixel data to exceed a predetermined number, the sub-pixel data separated by the farthest distance can be removed.
較佳地，點雲算圖方法可進一步設定子像素資料包含種子像素，影格原稿依據種子像素相隔設定距離的範圍中所包含的子像素資料來輸出影格資料，影格資料包含顯示顏色及顯示距離。 Preferably, the point cloud calculation method can further set the sub-pixel data to include seed pixels, and the frame manuscript outputs the frame data according to the sub-pixel data included in the range of the set distance between the seed pixels, and the frame data includes display color and display distance.
較佳地，可將相隔設定距離的範圍內的子像素資料所對應的顏色資料經過加權平均作為顯示顏色，對應的相隔距離經過加權平均作為顯示距離。 Preferably, the weighted average of the color data corresponding to the sub-pixel data within the range of the set distance can be used as the display color, and the weighted average of the corresponding distances can be used as the display distance.
較佳地，子像素資料還可包含終像素資料，當進行影格原稿的更新使子像素資料的數量超過預定數量時，將相隔距離最遠的子像素資料合併至終像素資料中。 Preferably, the sub-pixel data may also include final pixel data, and when the frame manuscript is updated so that the number of sub-pixel data exceeds a predetermined number, the sub-pixel data that are farthest apart are merged into the final pixel data.
較佳地，終像素資料可包含顏色和、距離和及權重和，顏色和為子像素資料所對應的顏色資料經由權重值及距離函數轉換後的總和，距離和為子像素資料對應的相隔距離經由權重值及距離函數轉換後的總和，權重和為子像素資料對應的權重值經由距離函數轉換後的總和。 Preferably, the final pixel data can include color sum, distance sum and weight sum, the color sum is the sum of the color data corresponding to the sub-pixel data converted by weight value and distance function, and the distance sum is the distance corresponding to the sub-pixel data The sum converted by the weight value and the distance function, the weight sum is the sum converted by the weight value corresponding to the sub-pixel data by the distance function.
較佳地，點雲算圖方法可進一步設定子像素資料包含種子像素，影格原稿依據種子像素相隔設定距離的範圍中所包含的子像素資料來輸出影格資料，影格資料包含顯示顏色及顯示距離。 Preferably, the point cloud calculation method can further set the sub-pixel data to include seed pixels, and the frame manuscript outputs the frame data according to the sub-pixel data included in the range of the set distance between the seed pixels, and the frame data includes display color and display distance.
較佳地，當設定距離的範圍內不包含終像素資料，可將相隔設定距離的範圍內的子像素資料所對應的顏色資料經過加權平均作為顯示顏色，對應的相隔距離經過加權平均作為顯示距離。 Preferably, when the final pixel data is not included within the range of the set distance, the color data corresponding to the sub-pixel data within the range of the set distance can be weighted and averaged as the display color, and the corresponding distance can be weighted and averaged as the display distance .
較佳地，當設定距離的範圍內包含終像素資料，可將相隔設定距離的範圍內的子像素資料併入終像素資料，再以終像素資料的顏色和除以權重和作為顯示顏色，距離和除以權重和作為顯示距離。 Preferably, when the final pixel data is included within the range of the set distance, the sub-pixel data within the range of the set distance can be merged into the final pixel data, and then the color sum of the final pixel data is divided by the weight sum as the display color, and the distance and divided by the weight sum as the display distance.
根據本發明之另一目的，提出一種點雲算圖方法，藉由處理器存取記憶體中的點雲資料，執行計算程序以更新影格原稿，影格原稿所呈現的投影平面對應於觀察中心，點雲算圖方法包含以下步驟：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含終像素資料；存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標；將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定；將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，將子像素合併至終像素資料中；以及藉由點雲資料完成影格原稿當中各像素的更新。 According to another object of the present invention, a point cloud calculation method is proposed. The processor accesses the point cloud data in the memory and executes the calculation program to update the original frame. The projection plane presented by the original frame corresponds to the observation center. The point cloud calculation method includes the following steps: setting the frame original, the frame original contains a pixel matrix composed of a plurality of pixels, and each pixel contains the final pixel data; accessing the position data and color data of the point cloud data, calculating the position data and observing the center The distance between the distance and the frame coordinates of the position data projected on the frame original; the frame coordinates are converted into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the frame original; the color data of the point cloud data, The distance and weight value are converted into sub-pixel data of each pixel, and the sub-pixels are merged into the final pixel data; and the update of each pixel in the original image frame is completed through the point cloud data.
較佳地，終像素資料可包含顏色和、距離和及權重和，顏色和為子像素資料所對應的顏色資料經由權重值及距離函數轉換後的總和，距離和為子像素資料對應的相隔距離經由權重值及距離函數轉換後的總和，權重和為子像素資料對應的權重值經由距離函數轉換後的總和。 Preferably, the final pixel data can include color sum, distance sum and weight sum, the color sum is the sum of the color data corresponding to the sub-pixel data converted by weight value and distance function, and the distance sum is the distance corresponding to the sub-pixel data The sum converted by the weight value and the distance function, the weight sum is the sum converted by the weight value corresponding to the sub-pixel data by the distance function.
較佳地，點雲算圖方法可進一步設定終像素資料包含種子像素，影格原稿依據種子像素來輸出影格資料，影格資料包含顯示顏色及顯示距離。 Preferably, the point cloud calculation method can further set the final pixel data to include seed pixels, and the frame manuscript outputs frame data according to the seed pixels, and the frame data includes display color and display distance.
較佳地，顯示顏色為終像素資料的顏色和除以權重和，顯示距離為距離和除以權重和。 Preferably, the display color is the color sum of the final pixel data divided by the weight sum, and the display distance is the distance sum divided by the weight sum.
承上所述，依本發明之點雲算圖方法，其可具有一或多個下述優點： As mentioned above, according to the point cloud calculation method of the present invention, it can have one or more of the following advantages:
(1)此點雲算圖方法能將點雲資料轉換成各個像素的子像素資料，通過距離排列及權重計算，使得每個像素的輸出結果更加接近實際場景，提升顯示品質，避免畫面產生顆粒、波紋等瑕疵問題。 (1) This point cloud calculation method can convert point cloud data into sub-pixel data of each pixel. Through distance arrangement and weight calculation, the output result of each pixel is closer to the actual scene, which improves the display quality and avoids grains in the picture. , ripples and other flaws.
(2)此點雲算圖方法能夠通過種子像素的設置，將預定範圍內的子像素加權平均，使得每個像素能呈現最適當的顯示資訊，提升轉換的正確性。 (2) This point cloud calculation method can weight and average the sub-pixels within a predetermined range through the setting of seed pixels, so that each pixel can present the most appropriate display information and improve the accuracy of conversion.
(3)此點雲算圖方法能將點雲資料轉換成投影平面上的顯示資料，提供影視、動畫相關的應用，提升操作的多樣性。 (3) This point cloud calculation method can convert point cloud data into display data on the projection plane, provide applications related to film and television, animation, and improve the diversity of operations.
11,51:點雲資料 11,51: point cloud data
12,52:子像素資料 12,52: sub-pixel data
13,23,33,43,53,63:影格原稿 13,23,33,43,53,63: frame original
24,44,64:影格資料 24,44,64: frame data
111,511:位置資料 111,511: location data
112,512:顏色資料 112,512: color data
121,521:顏色資料 121,521: color information
122,522:相隔距離 122,522: distance apart
123,523:權重值 123,523: weight value
241,441,641:顯示顏色 241,441,641: display color
242,442,642:顯示距離 242,442,642: display distance
P1~PN:第一子像素資料~第N子像素資料 P1~PN: first sub-pixel data ~ Nth sub-pixel data
PF:終像素資料 PF: final pixel data
PS:種子像素 PS: seed pixel
S11~S15,S21~S26,S31~S35,S41~S46,S51~S55,S61~S66:步驟 S11~S15, S21~S26, S31~S35, S41~S46, S51~S55, S61~S66: steps
為使本發明之技術特徵、內容與優點及其所能達成之功效更為顯而易見，茲將本發明配合附圖，並以實施例之表達形式詳細說明如下：第1A圖及第1B圖係為本發明第一實施例之點雲算圖方法之示意圖。 In order to make the technical features, content and advantages of the present invention and the effects that can be achieved more obvious, the present invention is hereby combined with the accompanying drawings, and is described in detail as follows in the expression form of the embodiment: Fig. 1A and Fig. 1B are A schematic diagram of the point cloud calculation method according to the first embodiment of the present invention.
第2A圖及第2B圖係為本發明第二實施例之點雲算圖方法之示意圖。 Fig. 2A and Fig. 2B are schematic diagrams of the point cloud computing method according to the second embodiment of the present invention.
第3A圖及第3B圖係為本發明第三實施例之點雲算圖方法之示意圖。 Fig. 3A and Fig. 3B are schematic diagrams of the point cloud computing method according to the third embodiment of the present invention.
第4A圖及第4B圖係為本發明第四實施例之點雲算圖方法之示意圖。 Fig. 4A and Fig. 4B are schematic diagrams of the point cloud computing method according to the fourth embodiment of the present invention.
第5A圖及第5B圖係為本發明第五實施例之點雲算圖方法之示意圖。 Fig. 5A and Fig. 5B are schematic diagrams of the point cloud computing method according to the fifth embodiment of the present invention.
第6A圖及第6B圖係為本發明第六實施例之點雲算圖方法之示意圖。 Fig. 6A and Fig. 6B are schematic diagrams of the point cloud computing method according to the sixth embodiment of the present invention.
第7A圖及第7B圖係為本發明實施例之點雲算圖方法與現有方法比較之示意圖。 Fig. 7A and Fig. 7B are schematic diagrams comparing the point cloud computing method of the embodiment of the present invention with the existing method.
為利貴審查委員瞭解本發明之技術特徵、內容與優點及其所能達成之功效，茲將本發明配合附圖，並以實施例之表達形式詳細說明如下，而其中所使用之圖式，其主旨僅為示意及輔助說明書之用，未必為本發明實施後之真實比例與精準配置，故不應就所附之圖式的比例與配置關係解讀、侷限本發明於實際實施上的權利範圍，合先敘明。 In order for the Ligui Examiner to understand the technical features, content and advantages of the present invention and the effects it can achieve, the present invention is hereby combined with the accompanying drawings and described in detail in the form of an embodiment as follows, and the drawings used therein, its The subject matter is only for illustration and auxiliary instructions, and not necessarily the true proportion and precise configuration of the present invention after implementation, so it should not be interpreted based on the proportion and configuration relationship of the attached drawings, and limit the scope of rights of the present invention in actual implementation. Together first describe.
請參閱第1A圖及第1B圖，其係為本發明第一實施例之點雲算圖方法之示意圖。其中第1A圖為點雲算圖方法之流程圖，第1B圖為點雲算圖方法之方塊圖。點雲是呈現空間座標中立體形狀、物件或場景的資料集，通過3D掃描、光達、雷射測距掃描、影像軟體推算等方式建構點雲的資料。點雲的檔案中可包含標頭區及資料區，標頭區包含資料型態、類別標籤等資訊，資料區則為空間中各個點資料的集合，點資料可包含位置座標(例如XYZ三軸座標值)、顏色資料(例如RGB資料)或者其他如反射強度、法向等資訊。當三維的點雲檔案要運用到影視、動畫等二維的顯示格式時，需要通過轉換的過程，即點雲算圖的流程，才能支援各種影像或畫面的應用。 Please refer to FIG. 1A and FIG. 1B , which are schematic diagrams of the point cloud calculation method according to the first embodiment of the present invention. Figure 1A is a flow chart of the point cloud calculation method, and Figure 1B is a block diagram of the point cloud calculation method. A point cloud is a data set that presents a three-dimensional shape, object, or scene in spatial coordinates. The point cloud data is constructed by means of 3D scanning, lidar, laser ranging scanning, and image software calculations. A point cloud file can include a header area and a data area. The header area contains information such as data type, category label, etc., and the data area is a collection of point data in the space. The point data can include position coordinates (such as XYZ three-axis coordinate value), color data (such as RGB data) or other information such as reflection strength, normal direction, etc. When 3D point cloud files are to be applied to 2D display formats such as film and television, animation, etc., a conversion process, that is, the process of point cloud calculation, is required to support the application of various images or screens.
點雲資料在經由上述掃描或推算方式取得後，可儲存於記憶體中，這裡所述的記憶體包含電腦的儲存裝置、伺服器的資料庫或是各種雲端資料庫中，當需進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第1A圖所示，點雲算圖的方法包含以下步驟(S11~S15)： The point cloud data can be stored in the memory after the above-mentioned scanning or calculation methods are obtained. The memory mentioned here includes the storage device of the computer, the database of the server or various cloud databases. When calculating the map, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program for calculating the map. As shown in Figure 1A, the method for calculating the point cloud map includes the following steps (S11~S15):
步驟S11：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含預定數量的子像素資料。影格原稿可為二維的像素矩陣，其矩陣呈現的平面為對應於觀察中心的投影平面，觀察中心(或者稱為相機中心) 是指由設定的視角或觀察位置來轉換點雲資料時的虛擬中心點。在影格原稿的各個像素當中，可以包含預定數量的子像素資料，子像素資料視為組成各個像素的多層資料，由子像素資料的內容決定各個像素的資料內容。 Step S11 : Setting a frame document, the frame document includes a pixel matrix composed of a plurality of pixels, and each pixel includes a predetermined number of sub-pixel data. The frame manuscript can be a two-dimensional pixel matrix, and the plane presented by the matrix is the projection plane corresponding to the observation center, the observation center (or called the camera center) It refers to the virtual center point when converting the point cloud data from the set angle of view or observation position. Each pixel of the frame original may contain a predetermined number of sub-pixel data. The sub-pixel data is regarded as multi-layer data composing each pixel, and the data content of each pixel is determined by the content of the sub-pixel data.
步驟S12：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。請參閱第1B圖，點雲資料11包含位置資料111及顏色資料112，當進行點雲算圖時，會計算位置資料111與觀察中心的相隔距離122，即點雲資料11所代表的點與觀察中心的距離。另外，還計算點雲資料11投影於影格原稿13上的影格座標，即點雲資料11所代表的點於呈現的平面上的投影位置。 Step S12: accessing the position data and color data of the point cloud data, calculating the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original. Please refer to Fig. 1B, the point cloud data 11 includes position data 111 and color data 112. When performing point cloud calculation, the distance 122 between the position data 111 and the observation center will be calculated, that is, the points represented by the point cloud data 11 and the The distance from the observation center. In addition, the frame coordinates of the point cloud data 11 projected on the frame manuscript 13 are also calculated, that is, the projection positions of the points represented by the point cloud data 11 on the presented plane.
步驟S13：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。一般而言，在點雲算圖時會取相隔距離122最近的點，將其顏色資料121作為該像素所呈現的顏色，但這樣的輸出方式往往使得輸出畫面具有顆粒感，擬真度不足，因此進一步設定權重值123，權重值123依據影格座標與影格原稿13中的像素座標之間的距離來決定，例如以影格座標減去像素座標後的絕對值來決定權重。由於投影的影格座標與影格原稿13當中各個像素的距離有所不同，將影格座標與該像素座標之間的距離越近的權重值123設定越高，讓越靠近像素座標的資訊具有較高的權重，以進行下一步的流程。 Step S13: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame. Generally speaking, when calculating a point cloud image, the point with the closest distance 122 will be taken, and its color data 121 will be used as the color presented by the pixel. However, such an output method often makes the output picture grainy, and the degree of fidelity is insufficient. Therefore, a weight value 123 is further set. The weight value 123 is determined according to the distance between the frame coordinates and the pixel coordinates in the frame original 13 , for example, the weight is determined by the absolute value of the frame coordinates minus the pixel coordinates. Since the distance between the projected frame coordinates and each pixel in the frame original 13 is different, the closer the distance between the frame coordinates and the pixel coordinates is, the higher the weight value 123 is set, so that information closer to the pixel coordinates has a higher weight. weights for the next step.
步驟S14：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，***原本的子像素資料當中並依據相隔距離由低至高的順序進行排列。當計算出點雲資料11的相隔距離122及權重值123後，將原本的點雲資料11轉換為子像素資料12，子像素資料12包含顏色資料121、相隔距離122 及權重值123。在影格原稿13當中，各個像素都會包含預定數量的子像素列，在本實施例中，子像素列包含第一子像素資料P1、第二子像素資料P2至第N子像素資料PN，這些子像素資料依據與觀察中心的距離遠近，由近至遠的順序進行排列，第一子像素資料P1所對應的點雲資料11為距離觀察中心最近的點。 Step S14: Convert the color data, separation distance and weight value of the point cloud data into sub-pixel data of each pixel, insert into the original sub-pixel data and arrange according to the order of separation distance from low to high. After calculating the separation distance 122 and weight value 123 of the point cloud data 11, the original point cloud data 11 is converted into sub-pixel data 12, and the sub-pixel data 12 includes color data 121, separation distance 122 and a weight value of 123. In the frame manuscript 13, each pixel will include a predetermined number of sub-pixel columns. In this embodiment, the sub-pixel columns include the first sub-pixel data P1, the second sub-pixel data P2 to the Nth sub-pixel data PN, these sub-pixel data The pixel data are arranged according to the distance from the observation center, from near to far. The point cloud data 11 corresponding to the first sub-pixel data P1 is the point closest to the observation center.
點雲資料11經由轉換形成子像素資料12後，將子像素資料12***於子像素列中，即第一子像素資料P1、第二子像素資料P2至第N子像素資料PN當中，依據相隔距離122由低至高進行排列後，形成新的子像素列。 After the point cloud data 11 is converted into sub-pixel data 12, the sub-pixel data 12 is inserted into the sub-pixel row, that is, the first sub-pixel data P1, the second sub-pixel data P2 to the Nth sub-pixel data PN, according to the distance After the distance 122 is arranged from low to high, a new sub-pixel row is formed.
步驟S15：藉由點雲資料完成影格原稿當中各像素的更新。點雲資料11所具備的資訊可以對影格原稿13中的各個像素進行更新，也就是對子像素列進行更新，然而，影格原稿13當中每個像素所具有的子像素資料數量有一定限制，當更新過程中，子像素資料數量超過原本的預定數量時，需要對超過的子像素資料進行處理。在本實施例中，當***的子像素資料12使得子像素列的數量超過預定數量時，移除最後一筆子像素資料，也就是將相隔距離最遠的子像素資料移除。 Step S15: Complete the updating of each pixel in the original frame by using the point cloud data. The information contained in the point cloud data 11 can update each pixel in the frame manuscript 13, that is, update the sub-pixel row. However, the number of sub-pixel data of each pixel in the frame manuscript 13 has a certain limit. During the update process, when the number of sub-pixel data exceeds the originally predetermined number, the excess sub-pixel data needs to be processed. In this embodiment, when the inserted sub-pixel data 12 causes the number of sub-pixel rows to exceed a predetermined number, the last piece of sub-pixel data is removed, that is, the sub-pixel data with the farthest distance is removed.
持續上述點雲算圖流程，將各個點雲資料更新影格原稿當中各個像素的子像素資料，當完成後，形成完整的影格原稿資料，後續可以依據此影格原稿資料來輸出所需的影格資料。 Continuing the above point cloud calculation process, each point cloud data is updated to the sub-pixel data of each pixel in the frame original. After completion, a complete frame original data is formed, and the required frame data can be output based on this frame original data.
請參閱第2A圖及第2B圖，其係為本發明第二實施例之點雲算圖方法之示意圖。其中第2A圖為點雲算圖方法之流程圖，第2B圖為種子像素之示意圖。進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第2A圖所示，點雲算圖的方法包含以下步驟(S21~S26)： Please refer to FIG. 2A and FIG. 2B, which are schematic diagrams of the point cloud calculation method according to the second embodiment of the present invention. Figure 2A is a flow chart of the point cloud calculation method, and Figure 2B is a schematic diagram of seed pixels. When performing point cloud calculation, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program of the calculation. As shown in FIG. 2A, the method for point cloud calculation includes the following steps (S21~ S26):
步驟S21：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含預定數量的子像素資料。 Step S21 : Setting a frame document, the frame document includes a pixel matrix composed of a plurality of pixels, and each pixel includes a predetermined number of sub-pixel data.
步驟S22：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。 Step S22: Access the position data and color data of the point cloud data, calculate the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original.
步驟S23：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。 Step S23: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame.
步驟S24：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，***原本的子像素資料當中並依據相隔距離由低至高的順序進行排列。 Step S24: Convert the color data, separation distance and weight value of the point cloud data into sub-pixel data of each pixel, insert into the original sub-pixel data and arrange according to the order of separation distance from low to high.
步驟S25：藉由點雲資料完成影格原稿當中各像素的更新。 Step S25: Complete the updating of each pixel in the original frame by using the point cloud data.
上述步驟S21至步驟S25的內容與第一實施例的內容類似，詳細說明請參閱前述實施例，相同內容在此不再重複描述。與第一實施例不同的是，在本實施例中，進一步加入了種子像素的設定以及輸出影格資料的步驟，其內容接續說明如下。 The content of the above step S21 to step S25 is similar to the content of the first embodiment. Please refer to the previous embodiment for details, and the same content will not be described again here. The difference from the first embodiment is that in this embodiment, the steps of setting the subpixels and outputting the frame data are further added, and the contents are described as follows.
步驟S26：設定子像素資料包含種子像素，影格原稿依據種子像素相隔設定距離的範圍中所包含的子像素資料來輸出影格資料，影格資料包含顯示顏色及顯示距離。請參閱第2B圖，第2B圖為種子像素的示意圖，如圖所示，影格原稿23當中包含了複數個像素形成的像素矩陣，每一個像素當中具有預定數量的子像素資料，例如子像素列包含第一子像素資料P1、第二子像素資料P2至第N子像素資料PN。在本實施例中，設定子像素資料當中第n個子像素資料為種子像素PS，n可為1到N中的任一數值。當設定種子像素PS後，再將種子像素
PS相隔設定距離的範圍內所包含的子像素資料輸出，作為實際的影格資料24，此影格資料24當中包含顯示顏色241及顯示距離242的資訊。
Step S26: Set the sub-pixel data to include sub-pixels, and the frame manuscript outputs the frame data according to the sub-pixel data included in the range of the set distance between the sub-pixels, and the frame data includes display colors and display distances. Please refer to FIG. 2B. FIG. 2B is a schematic diagram of a seed pixel. As shown in the figure, the frame manuscript 23 includes a pixel matrix formed by a plurality of pixels, and each pixel has a predetermined number of sub-pixel data, such as sub-pixel rows. It includes the first sub-pixel data P1, the second sub-pixel data P2 to the Nth sub-pixel data PN. In this embodiment, the nth sub-pixel data among the sub-pixel data is set as the sub-pixel PS, and n can be any value from 1 to N. After setting the seed pixel PS, the seed pixel
The sub-pixel data included in the range of the set distance apart from the PS is output as the actual frame data 24 , and the frame data 24 includes the information of the
當進行輸出時，可將相隔設定距離的範圍內的子像素資料所對應的顏色資料經過加權平均作為顯示顏色241，對應的相隔距離經過加權平均作為顯示距離242。
When outputting, the weighted average of the color data corresponding to the sub-pixel data within the range of the set distance can be used as the
請參閱第3A圖及第3B圖，其係為本發明第三實施例之點雲算圖方法之示意圖。其中第3A圖為點雲算圖方法之流程圖，第3B圖為影格原稿之示意圖。進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第3A圖所示，點雲算圖的方法包含以下步驟(S31~S35)： Please refer to FIG. 3A and FIG. 3B , which are schematic diagrams of the point cloud calculation method according to the third embodiment of the present invention. Figure 3A is a flowchart of the point cloud calculation method, and Figure 3B is a schematic diagram of a frame manuscript. When performing point cloud calculation, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program of the calculation. As shown in FIG. 3A, the method for point cloud calculation includes the following steps (S31~ S35):
步驟S31：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含預定數量的子像素資料，子像素資料還可包含終像素資料。請同時參閱第3B圖，第3B圖為本實施例之影格原稿的示意圖，如圖所示，影格原稿33包含複數個像素形成的像素矩陣，每一個像素當中具有預定數量的子像素資料，例如子像素列包含第一子像素資料P1、第二子像素資料P2至第N子像素資料PN。在本實施例中，子像素列還包含一個終像素資料PF，終像素資料PF可合併多個子像素資料，避免更新子像素資料時，產生過多的子像素資料。 Step S31 : setting the frame original, the frame original includes a pixel matrix composed of a plurality of pixels, each pixel includes a predetermined number of sub-pixel data, and the sub-pixel data may also include final pixel data. Please refer to Fig. 3B at the same time. Fig. 3B is a schematic diagram of the frame manuscript of this embodiment. As shown in the figure, the frame manuscript 33 includes a pixel matrix formed by a plurality of pixels, and each pixel has a predetermined number of sub-pixel data, for example The sub-pixel row includes first sub-pixel data P1, second sub-pixel data P2 to Nth sub-pixel data PN. In this embodiment, the sub-pixel row further includes a final pixel data PF, and the final pixel data PF can combine multiple sub-pixel data to avoid generating too much sub-pixel data when updating the sub-pixel data.
步驟S32：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。 Step S32: Access the position data and color data of the point cloud data, calculate the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original.
步驟S33：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。 Step S33: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame.
步驟S34：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，***原本的子像素資料當中並依據相隔距離由低至高的順 序進行排列。上述步驟S32至步驟S34的內容與第一實施例的內容類似，詳細說明請參閱前述實施例，相同內容在此不再重複描述。 Step S34: Convert the color data, separation distance, and weight value of the point cloud data into sub-pixel data of each pixel, insert them into the original sub-pixel data, and calculate them according to the order of separation distance from low to high. sorted in order. The content of the above step S32 to step S34 is similar to the content of the first embodiment. Please refer to the previous embodiment for details, and the same content will not be described again here.
步驟S35：藉由點雲資料完成影格原稿當中各像素的更新，當子像素資料的數量超過預定數量時，將相隔距離最遠的子像素資料合併至終像素資料中。如前述實施例所述，當進行子像素資料的更新時，會依據相隔距離重新對子像素資料進行排序，但更新後的子像素資料數量可能超過原本設定的數量，此時為保持資料的完整性，進一步將移除的這些子像素資料併入於終像素資料當中。 Step S35: Complete the updating of each pixel in the original frame by using the point cloud data, and merge the sub-pixel data with the farthest distance into the final pixel data when the number of sub-pixel data exceeds a predetermined number. As described in the foregoing embodiments, when sub-pixel data is updated, the sub-pixel data will be re-sorted according to the distance, but the number of updated sub-pixel data may exceed the original set number. At this time, to maintain the integrity of the data The removed sub-pixel data are further merged into the final pixel data.
在本實施例中，終像素資料可包含顏色和、距離和及權重和，顏色和為子像素資料所對應的顏色資料經由權重值及距離函數轉換後的總和，距離和為子像素資料對應的相隔距離經由權重值及距離函數轉換後的總和，權重和為子像素資料對應的權重值經由距離函數轉換後的總和。顏色和、距離和及權重和可以下列公式(1)至公式(3)進行計算。 In this embodiment, the final pixel data can include color sum, distance sum, and weight sum. The color sum is the sum of the color data corresponding to the sub-pixel data converted by the weight value and distance function, and the distance sum is the sum of the sub-pixel data. The distance is the sum converted by the weight value and the distance function, and the weight sum is the sum converted by the weight value corresponding to the sub-pixel data by the distance function. The color sum, distance sum, and weight sum can be calculated by the following formulas (1) to (3).
顏色和+=顏色資料×權重值×距離函數 (1) Color sum += color data × weight value × distance function (1)
距離和+=相隔距離×權重值×距離函數 (2) Distance sum += separation distance × weight value × distance function (2)
權重和+=權重值×距離函數 (3) Weight sum += weight value × distance function (3)
上述公式是指顏色和的增量為顏色資料×權重值×距離函數，距離和的增量為相隔距離×權重值×距離函數，權重和的增量為權重值×距離函數，其中距離函數是子像素距離的衰減函數。當取得上述增量後，子像素資料可通過下列公式(4)至公式(6)進行合併。 The above formula means that the increment of color sum is color data×weight value×distance function, the increment of distance sum is separation distance×weight value×distance function, the increment of weight sum is weight value×distance function, and the distance function is Decay function for subpixel distance. After the above increments are obtained, the sub-pixel data can be combined through the following formulas (4) to (6).
新顏色和=原顏色和+顏色和增量 (4) new color sum = original color sum + color sum increment (4)
新距離和=原距離和+距離和增量 (5) new distance sum = original distance sum + distance sum increment (5)
新權重和=原權重和+權重和增量 (6) new weight sum = old weight sum + weight sum delta (6)
對於常用的二進位浮點數算術標準(IEEE 754)的雙精度浮點數，能表示的正數值範圍通常是2-1074到21024(不含21024本身)，一般來說這個範圍已經非常足夠使用了。但衰減函數通常是指數函數或倒數多項式函數，例如距離函數是每5cm函數值就變為一半的指數衰減，如公式(7)所示。 For the double-precision floating-point numbers of the commonly used binary floating-point number arithmetic standard (IEEE 754), the positive value range that can be represented is usually 2 -1074 to 2 1024 (excluding 2 1024 itself), generally speaking, this range is very Enough to use. However, the attenuation function is usually an exponential function or a reciprocal polynomial function. For example, the distance function is an exponential attenuation in which the function value becomes half every 5 cm, as shown in formula (7).
此時由於雙精度浮點數的最小正數值2-1074限制，距離在54公尺以外的點，其距離函數會遇到浮點數下溢(underflow)為0。也就是說使用這樣的距離函數只能處理大約半徑50公尺內的點雲資料，為解決此問題，可利用對數壓縮距離和其他數值，以顏色和為例，將距離函數的公式(7)兩邊取對數，得到公式(8)，並將其帶入原顏色和公式而得到公式(9)。 At this time, due to the limit of the minimum positive value 2 -1074 of the double-precision floating-point number, the point whose distance is beyond 54 meters, the distance function will encounter the floating-point number underflow (underflow) to be 0. In other words, using such a distance function can only process point cloud data within a radius of about 50 meters. To solve this problem, logarithmic compression of distance and other values can be used. Taking the color sum as an example, the distance function formula (7) Taking the logarithm on both sides gives formula (8), which is brought into the original color sum formula to get formula (9).
距離和增量及權重和增量也可依據同樣方式計算，由於公式中都沒有原本距離函數公式中容易遇到浮點數下溢的項，因此可僅計算所有變數的對數值，直到輸出影像後再轉換回一般變數即可。由上述方式可利用對數壓縮的方式來取得更大的表示範圍。 The distance and increment and the weight and increment can also be calculated in the same way. Since there are no items in the formula that are prone to floating-point number underflow in the original distance function formula, only the logarithmic values of all variables can be calculated until the output image Then convert back to normal variables. In the above way, logarithmic compression can be used to obtain a larger representation range.
另外，上述的對數運算可將冪運算轉換為乘法，將乘法轉換為加法，但在遇到加法時會較為複雜，將加法轉換為log 2 sum表示，請參閱公式(10)。 In addition, the above-mentioned logarithmic operation can convert exponentiation into multiplication and multiplication into addition, but it will be more complicated when encountering addition, and the addition can be converted into log 2 sum representation, please refer to formula (10).
log2 sum(a,b)=max(a,b)+log2(1+0.5|a-b|) (10) log 2 sum ( a,b )=max(a,b)+log 2 (1+0.5 | a - b | ) (10)
不過在使用上述形式時，在∞=∞+∞時會有∞-∞的錯誤，因此將上述公式稍做變形，明確算出當a與b相等時的情況，如公式(11)所示。 However, when using the above form, there will be an ∞-∞ error when ∞=∞+∞, so the above formula is slightly deformed to clearly calculate the situation when a and b are equal, as shown in formula (11).
log2 sum(a,b)=max(a,b)+1[a=b]+log2(1+0.5|a-b|)[a≠b] (11) log 2 sum ( a,b )=max( a,b )+1[ a = b ]+log 2 (1+0.5 | a - b | )[ a ≠ b ] (11)
上述公式是通過艾佛森括號(Iverson Bracket)的方式，括號內的條件滿足則為1，不滿足則為0。此外，由於log2(1+0.5|a-b|)數值較小，可視為修正項，以近似的指數函數加速計算，如公式(12)所示。 The above formula uses Iverson brackets (Iverson Bracket), and the conditions in the brackets are 1 if they are met, and 0 if they are not. In addition, since the value of log 2 (1+0.5 | a - b | ) is small, it can be regarded as a correction item, and the calculation can be accelerated with an approximate exponential function, as shown in formula (12).
請參閱第4A圖及第4B圖，其係為本發明第四實施例之點雲算圖方法之示意圖。其中第4A圖為點雲算圖方法之流程圖，第4B圖為種子像素之示意圖。進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第4A圖所示，點雲算圖的方法包含以下步驟(S41~S46)： Please refer to FIG. 4A and FIG. 4B, which are schematic diagrams of the point cloud calculation method according to the fourth embodiment of the present invention. Figure 4A is a flow chart of the point cloud calculation method, and Figure 4B is a schematic diagram of a seed pixel. When performing point cloud calculation, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program of the calculation. As shown in FIG. 4A, the method for point cloud calculation includes the following steps (S41~ S46):
步驟S41：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含預定數量的子像素資料，子像素資料還可包含終像素資料。 Step S41 : Set a frame document. The frame document includes a pixel matrix composed of a plurality of pixels, and each pixel includes a predetermined number of sub-pixel data, and the sub-pixel data may also include final pixel data.
步驟S42：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。 Step S42: Access the position data and color data of the point cloud data, calculate the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original.
步驟S43：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。 Step S43: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame.
步驟S44：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，***原本的子像素資料當中並依據相隔距離由低至高的順序進行排列。 Step S44: Convert the color data, separation distance and weight value of the point cloud data into sub-pixel data of each pixel, insert into the original sub-pixel data and arrange according to the order of separation distance from low to high.
步驟S45：藉由點雲資料完成影格原稿當中各像素的更新，當子像素資料的數量超過預定數量時，將相隔距離最遠的子像素資料合併至終像素資料中。 Step S45: Complete the updating of each pixel in the original frame by using the point cloud data, and merge the sub-pixel data with the farthest distance into the final pixel data when the number of sub-pixel data exceeds a predetermined number.
上述步驟S41至步驟S45的內容與第三實施例的內容類似，詳細說明請參閱前述實施例，相同內容在此不再重複描述。與第三實施例不同的是，在本實施例中，進一步加入了種子像素的設定以及輸出影格資料的步驟，其內容接續說明如下。 The content of the above step S41 to step S45 is similar to the content of the third embodiment. Please refer to the previous embodiment for details, and the same content will not be described again here. The difference from the third embodiment is that in this embodiment, the steps of setting the subpixels and outputting the frame data are further added, and the contents are described as follows.
步驟S46：設定子像素資料包含種子像素，影格原稿依據種子像素相隔設定距離的範圍中所包含的子像素資料來輸出影格資料，影格資料包含顯示顏色及顯示距離。請參閱第4B圖，第4B圖為種子像素的示意圖，如圖所示，影格原稿43當中包含了複數個像素形成的像素矩陣，每一個像素當中具有預定數量的子像素資料，例如子像素列包含第一子像素資料P1、第二子像素資料P2至第N子像素資料PN以及終像素資料PF。在本實施例中，設定子像素資料當中第n個子像素資料為種子像素PS，n可為1到N中的任一數值，也可設定終像素資料PF為種子像素PS。當設定種子像素PS後，再將種子像素PS相隔設定距離的範圍內所包含的子像素資料輸出，作為實際的影格資料44，此影格資料44當中包含顯示顏色441及顯示距離442的資訊。
Step S46: Set the sub-pixel data to include sub-pixels, and the frame manuscript outputs the frame data according to the sub-pixel data included in the range of the set distance between the sub-pixels, and the frame data includes display colors and display distances. Please refer to FIG. 4B. FIG. 4B is a schematic diagram of a seed pixel. As shown in the figure, the frame manuscript 43 includes a pixel matrix formed by a plurality of pixels, and each pixel has a predetermined number of sub-pixel data, such as sub-pixel columns. It includes the first sub-pixel data P1, the second sub-pixel data P2 to the Nth sub-pixel data PN and the final pixel data PF. In this embodiment, the nth sub-pixel data among the sub-pixel data is set as the sub-pixel PS, and n can be any value from 1 to N, and the final pixel data PF can also be set as the sub-pixel PS. After the sub-pixel PS is set, the sub-pixel data included in the range of the set distance between the sub-pixel PS is output as the actual frame data 44 . The frame data 44 includes
當設定距離的範圍內不包含終像素資料PF，可將相隔設定距離的範圍內的子像素資料所對應的顏色資料經過加權平均作為顯示顏色441，對應的相隔距離經過加權平均作為顯示距離442。這部分與前述實施例類似，但當設定距離的範圍內包含終像素資料PF時，即當終像素資料PF的相隔距離，也就是距離和除以權重和在設定距離的範圍內時，其輸出方式與前述實施例有所差異。
在本實施例中，當設定距離的範圍包含終像素資料PF，可將相隔設定距離的範圍內的子像素資料併入終像素資料PF，其合併方式與前述顏色和、距離和及權重和的合併方式相同，當資料都併入終像素資料PF後，再以終像素資料PF的顏色和除以權重和作為顯示顏色441，距離和除以權重和作為顯示距離442。
When the final pixel data PF is not included in the range of the set distance, the color data corresponding to the sub-pixel data within the range of the set distance can be weighted and averaged as the
請參閱第5A圖及第5B圖，其係為本發明第五實施例之點雲算圖方法之示意圖。其中第5A圖為點雲算圖方法之流程圖，第5B圖為影格原稿之示意圖。當需進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第5A圖所示，點雲算圖的方法包含以下步驟(S51~S55)： Please refer to FIG. 5A and FIG. 5B, which are schematic diagrams of the point cloud calculation method according to the fifth embodiment of the present invention. Figure 5A is a flow chart of the point cloud calculation method, and Figure 5B is a schematic diagram of a frame manuscript. When it is necessary to perform point cloud calculation, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program for calculation. As shown in Figure 5A, the method for point cloud calculation includes the following steps ( S51~S55):
步驟S51：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含終像素資料。在本實施例中，各個像素當中不再設置預定數量的子像素資料，而是直接設置終像素資料。終像素資料可包含顏色和、距離和及權重和的資料，其中，顏色和為子像素資料所對應的顏色資料經由權重值及距離函數轉換後的總和，距離和為子像素資料對應的相隔距離經由權重值及距離函數轉換後的總和，權重和為子像素資料對應的權重值經由距離函數轉換後的總和，距離函數可參照上述公式(1)至公式(3)之說明。 Step S51 : Setting the frame manuscript, the frame manuscript includes a pixel matrix composed of a plurality of pixels, and each pixel contains final pixel data. In this embodiment, instead of setting a predetermined number of sub-pixel data in each pixel, the final pixel data is directly set. The final pixel data can include color sum, distance sum and weight sum data, where the color sum is the sum of the color data corresponding to the sub-pixel data converted by the weight value and distance function, and the distance sum is the distance corresponding to the sub-pixel data The sum converted by the weight value and the distance function, the weight sum is the sum converted by the weight value corresponding to the sub-pixel data by the distance function, the distance function can refer to the description of the above formula (1) to formula (3).
步驟S52：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。請參閱第5B圖，點雲資料51包含位置資料511及顏色資料512，當進行點雲算圖時，會計算位置資料511與觀察中心的相隔距離522，即點雲資料51所代表的點與觀察中心的距離。另外，還計算點雲資料51投影於影格原稿53上的影格座標，即點雲資料51所代表的點於呈現的平面上的投影位置。 Step S52: Access the position data and color data of the point cloud data, calculate the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original. Please refer to FIG. 5B , the point cloud data 51 includes position data 511 and color data 512. When performing point cloud calculation, the distance 522 between the position data 511 and the observation center will be calculated, that is, the points represented by the point cloud data 51 and The distance from the observation center. In addition, the frame coordinates of the point cloud data 51 projected on the frame manuscript 53 are also calculated, that is, the projected positions of the points represented by the point cloud data 51 on the presented plane.
步驟S53：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。設定權重值523，權重值523依據影格座標與影格原稿53中的像素座標之間的距離來決定，例如以影格座標減去像素座標後的絕對值來決定權重。由於投影的影格座標與影格原稿53當中各個像素的距離有所不同，將影格座標與該像素座標之間的距離越近的權重值523設定越高，讓越靠近像素座標的資訊具有較高的權重。 Step S53: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame. The weight value 523 is set. The weight value 523 is determined according to the distance between the frame coordinates and the pixel coordinates in the frame original 53 , for example, the weight is determined by the absolute value obtained by subtracting the pixel coordinates from the frame coordinates. Since the distance between the projected frame coordinates and each pixel in the frame original 53 is different, the closer the distance between the frame coordinates and the pixel coordinates is, the higher the weight value 523 is set, so that information closer to the pixel coordinates has a higher weight. Weights.
步驟S54：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，子像素資料合併至終像素資料中。當計算出點雲資料51的相隔距離522及權重值523後，將原本的點雲資料51轉換為子像素資料52，子像素資料52包含顏色資料521、相隔距離522及權重值523。將子像素資料52合併至終像素資料PF中。 Step S54: Convert the color data, separation distance and weight value of the point cloud data into sub-pixel data of each pixel, and merge the sub-pixel data into the final pixel data. After the separation distance 522 and weight value 523 of the point cloud data 51 are calculated, the original point cloud data 51 is converted into sub-pixel data 52 , and the sub-pixel data 52 includes color data 521 , separation distance 522 and weight value 523 . Merge the sub-pixel data 52 into the final pixel data PF.
步驟S55：藉由點雲資料完成影格原稿當中各像素的更新。點雲資料51所具備的資訊轉換成子像素資料52後，持續合併至終像素資料PF當中，以對影格原稿53中的各個像素進行更新。 Step S55: Complete the update of each pixel in the original frame by using the point cloud data. After the information contained in the point cloud data 51 is converted into sub-pixel data 52 , it is continuously merged into the final pixel data PF to update each pixel in the original frame 53 .
請參閱第6A圖及第6B圖，其係為本發明第六實施例之點雲算圖方法之示意圖。其中第6A圖為點雲算圖方法之流程圖，第6B圖為種子像素之示意圖。當需進行點雲算圖時，通過電腦裝置中的處理器存取記憶體中的點雲資料，執行算圖的計算程序，如第6A圖所示，點雲算圖的方法包含以下步驟(S61~S66)： Please refer to FIG. 6A and FIG. 6B, which are schematic diagrams of the point cloud calculation method according to the sixth embodiment of the present invention. Figure 6A is a flow chart of the point cloud calculation method, and Figure 6B is a schematic diagram of a seed pixel. When it is necessary to perform point cloud calculation, the processor in the computer device accesses the point cloud data in the memory, and executes the calculation program for calculation. As shown in Figure 6A, the method for point cloud calculation includes the following steps ( S61~S66):
步驟S61：設定影格原稿，影格原稿包含複數個像素組成的像素矩陣，各像素分別包含終像素資料。 Step S61 : Setting the frame manuscript, the frame manuscript includes a pixel matrix composed of a plurality of pixels, and each pixel contains final pixel data.
步驟S62：存取點雲資料的位置資料及顏色資料，計算位置資料與觀察中心的相隔距離及位置資料投影於影格原稿上的影格座標。 Step S62: Access the position data and color data of the point cloud data, calculate the distance between the position data and the observation center and the frame coordinates of the position data projected on the frame original.
步驟S63：將影格座標轉換為權重值，權重值依據影格座標與影格原稿中的像素座標之間的距離來決定。 Step S63: Convert the frame coordinates into weight values, and the weight values are determined according to the distance between the frame coordinates and the pixel coordinates in the original frame.
步驟S64：將點雲資料的顏色資料、相隔距離及權重值轉換為各像素的子像素資料，子像素資料合併至終像素資料中。 Step S64: Convert the color data, separation distance and weight value of the point cloud data into sub-pixel data of each pixel, and merge the sub-pixel data into the final pixel data.
步驟S65：藉由點雲資料完成影格原稿當中各像素的更新。 Step S65: Complete the update of each pixel in the original frame by using the point cloud data.
上述步驟S61至步驟S65的內容與第五實施例的內容類似，詳細說明請參閱前述實施例，相同內容在此不再重複描述。與第五實施例不同的是，在本實施例中，進一步加入了種子像素的設定以及輸出影格資料的步驟，其內容接續說明如下。 The content of the above step S61 to step S65 is similar to the content of the fifth embodiment. Please refer to the previous embodiment for details, and the same content will not be described again here. The difference from the fifth embodiment is that in this embodiment, the steps of setting the subpixels and outputting the frame data are further added, and the contents are described as follows.
步驟S66：設定終像素資料包含種子像素，影格原稿依據種子像素來輸出影格資料，影格資料包含顯示顏色及顯示距離。請參閱第6B圖，第6B圖為種子像素的示意圖，如圖所示，影格原稿63當中包含了複數個像素形成的像素矩陣，每一個像素當中具有終像素資料PF，將終像素資料PF設定為種子像素PS，以種子像素PS的內容來將合併的子像素資料輸出，作為實際的影格資料64，此影格資料64當中包含顯示顏色641及顯示距離642的資訊。
Step S66 : Set the final pixel data to include seed pixels, and the frame manuscript outputs frame data according to the seed pixels, and the frame data includes display color and display distance. Please refer to FIG. 6B. FIG. 6B is a schematic diagram of a seed pixel. As shown in the figure, the frame manuscript 63 includes a pixel matrix formed by a plurality of pixels, each pixel has final pixel data PF, and the final pixel data PF is set The combined sub-pixel data is output as the sub-pixel PS with the content of the sub-pixel PS as the actual frame data 64 , and the frame data 64 includes the information of the
在本實施例中，顯示顏色641為終像素資料PF的顏色和除以權重和，顯示距離642為距離和除以權重和。
In this embodiment, the
請參閱第7A圖及第7B圖，其係為本發明實施例之點雲算圖方法與現有方法比較之示意圖。其中第7A圖為現有方法轉換後的顯示畫面，第7B圖為本實施例之點雲算圖方法所產生的顯示畫面。由第7A圖所示，現有的點雲算圖方法，最後產生的圖會具有雜點，造成顆粒感或波紋的情況，當應用於影視或動畫等場景時，無法產生擬真的效果，也難以接近人類視覺的感受畫面。然而， 經由本實施例的點雲算圖方法計算後，實際輸出的畫面可如第7B圖所示，通過前述權重值或種子像素等運算方式，明顯可改善顯示品質，讓畫面更為接近實際場景，取得更為高品質的呈現效果。 Please refer to FIG. 7A and FIG. 7B, which are schematic diagrams comparing the point cloud calculation method of the embodiment of the present invention with the existing method. Figure 7A is the display screen converted by the existing method, and Figure 7B is the display screen generated by the point cloud calculation method of this embodiment. As shown in Figure 7A, the existing point cloud calculation method, the final image will have noise points, resulting in graininess or ripples. When applied to scenes such as film and television or animation, it cannot produce realistic effects, nor can it It is difficult to approach the feeling picture of human vision. However, After calculation by the point cloud calculation method of this embodiment, the actual output picture can be shown in Figure 7B. Through the calculation methods such as the aforementioned weight value or seed pixel, the display quality can be obviously improved, and the picture can be closer to the actual scene. Achieve higher quality rendering.
以上所述僅為舉例性，而非為限制性者。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，均應包含於後附之申請專利範圍中。 The above descriptions are illustrative only, not restrictive. Any equivalent modification or change made without departing from the spirit and scope of the present invention shall be included in the scope of the appended patent application.
S11~S15:步驟 S11~S15: Steps
Claims (13)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW110144931A TWI782806B (en) | 2021-12-02 | 2021-12-02 | Point cloud rendering method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW110144931A TWI782806B (en) | 2021-12-02 | 2021-12-02 | Point cloud rendering method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI782806B true TWI782806B (en) | 2022-11-01 |
| TW202324307A TW202324307A (en) | 2023-06-16 |
Family
ID=85794358
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW110144931A TWI782806B (en) | 2021-12-02 | 2021-12-02 | Point cloud rendering method |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI782806B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103828359A (en) * | 2011-09-29 | 2014-05-28 | 杜比实验室特许公司 | Representation and coding of multi-view images using tapestry encoding |
| TW201937405A (en) * | 2018-02-26 | 2019-09-16 | 財團法人工業技術研究院 | System and method for object labeling |
| CN111133476A (en) * | 2017-09-18 | 2020-05-08 | 苹果公司 | Point cloud compression |
| TW202034215A (en) * | 2018-08-13 | 2020-09-16 | 英國帝國科技及醫學學院 | Mapping object instances using video data |
| TW202103153A (en) * | 2019-07-04 | 2021-01-16 | 大陸商浙江商湯科技開發有限公司 | Method and electronic device for a point cloud fusion, and computer storage medium thereof |
-
2021
- 2021-12-02 TW TW110144931A patent/TWI782806B/en active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103828359A (en) * | 2011-09-29 | 2014-05-28 | 杜比实验室特许公司 | Representation and coding of multi-view images using tapestry encoding |
| CN111133476A (en) * | 2017-09-18 | 2020-05-08 | 苹果公司 | Point cloud compression |
| TW201937405A (en) * | 2018-02-26 | 2019-09-16 | 財團法人工業技術研究院 | System and method for object labeling |
| TW202034215A (en) * | 2018-08-13 | 2020-09-16 | 英國帝國科技及醫學學院 | Mapping object instances using video data |
| TW202103153A (en) * | 2019-07-04 | 2021-01-16 | 大陸商浙江商湯科技開發有限公司 | Method and electronic device for a point cloud fusion, and computer storage medium thereof |
| TWI722638B (en) * | 2019-07-04 | 2021-03-21 | 大陸商浙江商湯科技開發有限公司 | Method and electronic device for a point cloud fusion, and computer storage medium thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202324307A (en) | 2023-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20230053462A1 (en) | Image rendering method and apparatus, device, medium, and computer program product | |
| US6229549B1 (en) | High-speed three-dimensional texture mapping systems and methods | |
| US11908107B2 (en) | Method and apparatus for presenting image for virtual reality device, device and non-transitory computer-readable storage medium | |
| JP2009080578A (en) | Multiview-data generating apparatus, method, and program | |
| JP2005228320A (en) | Method and apparatus for high-speed visualization of depth image-based three-dimensional graphic data, and program-recorded computer readable recording medium | |
| WO2018113502A1 (en) | Method for automatically generating grid and multi-level of detail of shader | |
| CN1617173A (en) | Map display apparatus | |
| WO2019041842A1 (en) | Image processing method and device, storage medium and computer device | |
| CN112184873B (en) | Fractal graph creation method, fractal graph creation device, electronic equipment and storage medium | |
| CN108022223A (en) | A kind of tone mapping method based on the processing fusion of logarithmic mapping function piecemeal | |
| WO2022063260A1 (en) | Rendering method and apparatus, and device | |
| CN113643414B (en) | Three-dimensional image generation method and device, electronic equipment and storage medium | |
| WO2023142607A1 (en) | Image rendering method and apparatus, and device and medium | |
| JP2023552538A (en) | Image processing methods and devices, electronic devices, storage media, and computer programs | |
| CN113781621A (en) | Three-dimensional reconstruction processing method, device, equipment and storage medium | |
| CN110634178A (en) | Three-dimensional scene refinement reconstruction method for digital museum | |
| CN111343444B (en) | Three-dimensional image generation method and device | |
| TWI439961B (en) | Conversion algorithm for voids generated after converting 2D images | |
| CN110544318B (en) | Massive model loading method based on scene resolution of display window | |
| CN111311720B (en) | Texture image processing method and device | |
| JP3971714B2 (en) | Virtual viewpoint image generation method, virtual viewpoint image generation apparatus, virtual viewpoint image generation program, and recording medium | |
| CN107369138B (en) | Image optimization display method based on high-order statistical model | |
| TWI782806B (en) | Point cloud rendering method | |
| US20120038785A1 (en) | Method for producing high resolution image | |
| US20200334890A1 (en) | Systems and Methods for Fast Texture Mapping of 3D Models |