ๆญฃๅœจๅŠ ่ฝฝๅ›พ็‰‡...
1:12 Y.Yin et al. basis vectors w1,w2 represent the p-axis and yp-axis,respectively.With p-axis and yp-axis, we further calculate the zp-axis as w3=wi x w2,to establish the principal-plane coordinate system (principal-frame for short).Here,w1,w2,w3 are described in human-frame.While in the principal-frame,we can represent zp-axis,yp-axis,zp-axis as the unit vector(1,0,0)T, (0,1,0)T,(0,0,1)T,respectively.By comparing w1,w2,w3 in human-frame and xp-axis, Up-axis,Zp-axis in principal-frame,we can get the rotation matrix Rhp,which transforms coordinates from human-frame to principal-frame,as shown in Eq.(5)and Eq.(6). 1 00 1 (5) 01 w 7 (6) With the rotation matrix Rhp,we then calculate the projection of each axis of human- frame in principal plane.For convenience,we use ui,iE[1,3]to represent Th-axis,Uh-axis, Zh-axis,respectively.For ui,its coordinates in the principal-frame is qi,where gi=Rhpui. Then,we get the projection v;of qi in the principal plane with M,i.e.,setting the coordinate value in Zp-axis to zero,as shown in Eq.(7). ใ€Œ100] vi=Mqi= 010 qi (7) 000 As shown in Fig.10(b),Fig.11(b),Fig.12(b),Fig.13(b)and Fig.14(b),we represent the projected axis of h-axis,yh-axis,zh-axis (of human-frame)in the principal plane with black,green,fuchsia dashed line,respectively. 4.2.2 Reference Plane Detection.In Fig.15,we show how to utilize the length of the projected axis to detect the reference coordinate plane.Intuitively,if the projection of axis wi has shortest length in the principal plane,it indicates that the coordinate plane perpendicular to wi has the highest probability of being parallel to the principal plane,and should be selected as the reference coordinate plane for contour calibration.In the following,we will provide the proof for this intuition. As shown in Fig.15,the plane ABCD and the plane EFGH intersect on line CD(i.e., EF).For simplification,we use Op to represent one of the three axes (i.e.,Zh-axis,yh axis,Zh-axis),the plane ABCD represents the corresponding plane perpendicular to O1P (i.e.,yh-Zh plane,Th-zh plane,Th-yh plane),while EFGH represents the principal plane.To obtain the projection of OP,we first extend the line OP to intersect with EFGH at P,then O1P'LCD,because O1P L ABCD.Besides,we make the line O1IโŠฅCD,then we get OIโŠฅCD and O1P'โŠฅCD,thus CDโŠฅโ–ณOIP'.From point O1,we make the line 010,where 01O1 IP'.Then,we obtain that 010LCD and O1OLIP/,thus 010L EFGH.Similarly,from point P,we make the line PO2,where PO2 LIP'and PO2//010.Therefore,002 is the projection of O1P in the principal plane,0102=01P].sin 01. n regard to01and0g,01+02=90°๏ผŒ2+0g=90°๏ผŒthus01=s,where 03 means the plane included angle between ABCD and EFGH.If 03(or 01)is equal to zero,then the plane ABCD is parallel to the plane EFGH,and ABCD will be selected as reference coordinate plane,in what the projected 2D contour is similar to that in principal plane in high degree. ACM Trans.Sensor Netw.,Vol.1,No.1,Article 1.Publication date:January 2019.1:12 Y. Yin et al. basis vectors ๐œ”1, ๐œ”2 represent the ๐‘ฅ๐‘-axis and ๐‘ฆ๐‘-axis, respectively. With ๐‘ฅ๐‘-axis and ๐‘ฆ๐‘-axis, we further calculate the ๐‘ง๐‘-axis as ๐œ”3 = ๐œ”1 × ๐œ”2, to establish the principal-plane coordinate system (principal-frame for short). Here, ๐œ”1, ๐œ”2, ๐œ”3 are described in human-frame. While in the principal-frame, we can represent ๐‘ฅ๐‘-axis, ๐‘ฆ๐‘-axis, ๐‘ง๐‘-axis as the unit vector (1, 0, 0)๐‘‡ , (0, 1, 0)๐‘‡ , (0, 0, 1)๐‘‡ , respectively. By comparing ๐œ”1, ๐œ”2, ๐œ”3 in human-frame and ๐‘ฅ๐‘-axis, ๐‘ฆ๐‘-axis, ๐‘ง๐‘-axis in principal-frame, we can get the rotation matrix ๐‘…โ„Ž๐‘, which transforms coordinates from human-frame to principal-frame, as shown in Eq. (5) and Eq. (6). โŽก โŽฃ 1 0 0 0 1 0 0 0 1 โŽค โŽฆ = ๐‘…โ„Ž๐‘ โŽก โŽฃ ๐œ” ๐‘‡ 1 ๐œ” ๐‘‡ 2 ๐œ” ๐‘‡ 3 โŽค โŽฆ (5) ๐‘…โ„Ž๐‘ = โŽก โŽฃ ๐œ” ๐‘‡ 1 ๐œ” ๐‘‡ 2 ๐œ” ๐‘‡ 3 โŽค โŽฆ โˆ’1 (6) With the rotation matrix ๐‘…โ„Ž๐‘, we then calculate the projection of each axis of human๏ฟพframe in principal plane. For convenience, we use ๐‘ข๐‘–, ๐‘– โˆˆ [1, 3] to represent ๐‘ฅโ„Ž-axis, ๐‘ฆโ„Ž-axis, ๐‘งโ„Ž-axis, respectively. For ๐‘ข๐‘–, its coordinates in the principal-frame is ๐‘ž๐‘–, where ๐‘ž๐‘– = ๐‘…โ„Ž๐‘๐‘ข๐‘–. Then, we get the projection ๐‘ฃ๐‘– of ๐‘ž๐‘– in the principal plane with ๐‘€, i.e., setting the coordinate value in ๐‘ง๐‘-axis to zero, as shown in Eq. (7). ๐‘ฃ๐‘– = ๐‘€ ๐‘ž๐‘– = โŽก โŽฃ 1 0 0 0 1 0 0 0 0 โŽค โŽฆ ๐‘ž๐‘– (7) As shown in Fig. 10(b), Fig. 11(b), Fig. 12(b), Fig. 13(b) and Fig. 14(b), we represent the projected axis of ๐‘ฅโ„Ž-axis, ๐‘ฆโ„Ž-axis, ๐‘งโ„Ž-axis (of human-frame) in the principal plane with black, green, fuchsia dashed line, respectively. 4.2.2 Reference Plane Detection. In Fig. 15, we show how to utilize the length of the projected axis to detect the reference coordinate plane. Intuitively, if the projection of axis ๐œ”๐‘– has shortest length in the principal plane, it indicates that the coordinate plane perpendicular to ๐œ”๐‘– has the highest probability of being parallel to the principal plane, and should be selected as the reference coordinate plane for contour calibration. In the following, we will provide the proof for this intuition. As shown in Fig. 15, the plane ๐ด๐ต๐ถ๐ท and the plane ๐ธ๐น ๐บ๐ป intersect on line ๐ถ๐ท (i.e., ๐ธ๐น). For simplification, we use โˆ’โˆ’โ†’๐‘‚1๐‘ƒ to represent one of the three axes (i.e., ๐‘ฅโ„Ž-axis, ๐‘ฆโ„Ž- axis, ๐‘งโ„Ž-axis), the plane ๐ด๐ต๐ถ๐ท represents the corresponding plane perpendicular to โˆ’โˆ’โ†’๐‘‚1๐‘ƒ (i.e., ๐‘ฆโ„Ž โˆ’ ๐‘งโ„Ž plane, ๐‘ฅโ„Ž โˆ’ ๐‘งโ„Ž plane, ๐‘ฅโ„Ž โˆ’ ๐‘ฆโ„Ž plane), while ๐ธ๐น ๐บ๐ป represents the principal plane. To obtain the projection of ๐‘‚1๐‘ƒ, we first extend the line ๐‘‚1๐‘ƒ to intersect with ๐ธ๐น ๐บ๐ป at ๐‘ƒ โ€ฒ , then ๐‘‚1๐‘ƒ โ€ฒ โŠฅ ๐ถ๐ท, because ๐‘‚1๐‘ƒ โŠฅ ๐ด๐ต๐ถ๐ท. Besides, we make the line ๐‘‚1๐ผ โŠฅ ๐ถ๐ท, then we get ๐‘‚1๐ผ โŠฅ ๐ถ๐ท and ๐‘‚1๐‘ƒ โ€ฒ โŠฅ ๐ถ๐ท, thus ๐ถ๐ท โŠฅ โ–ณ๐‘‚1๐ผ๐‘ƒโ€ฒ . From point ๐‘‚1, we make the line ๐‘‚1๐‘‚โ€ฒ 1 , where ๐‘‚1๐‘‚โ€ฒ 1 โŠฅ ๐ผ๐‘ƒโ€ฒ . Then, we obtain that ๐‘‚1๐‘‚โ€ฒ 1 โŠฅ ๐ถ๐ท and ๐‘‚1๐‘‚โ€ฒ 1 โŠฅ ๐ผ๐‘ƒโ€ฒ , thus ๐‘‚1๐‘‚โ€ฒ 1 โŠฅ ๐ธ๐น ๐บ๐ป. Similarly, from point ๐‘ƒ, we make the line ๐‘ƒ ๐‘‚2, where ๐‘ƒ ๐‘‚2 โŠฅ ๐ผ๐‘ƒโ€ฒ and ๐‘ƒ ๐‘‚2//๐‘‚1๐‘‚โ€ฒ 1 . Therefore, ๐‘‚โ€ฒ 1๐‘‚2 is the projection of ๐‘‚1๐‘ƒ in the principal plane, |๐‘‚โ€ฒ 1๐‘‚2| = |๐‘‚1๐‘ƒ| · sin ๐œƒ1. In regard to ๐œƒ1 and ๐œƒ3, ๐œƒ1+๐œƒ2 = 90โˆ˜ , ๐œƒ2+๐œƒ3 = 90โˆ˜ , thus ๐œƒ1 = ๐œƒ3, where ๐œƒ3 means the plane included angle between ๐ด๐ต๐ถ๐ท and ๐ธ๐น ๐บ๐ป. If ๐œƒ3 (or ๐œƒ1) is equal to zero, then the plane ๐ด๐ต๐ถ๐ท is parallel to the plane ๐ธ๐น ๐บ๐ป, and ๐ด๐ต๐ถ๐ท will be selected as reference coordinate plane, in what the projected 2D contour is similar to that in principal plane in high degree. ACM Trans. Sensor Netw., Vol. 1, No. 1, Article 1. Publication date: January 2019
<<ๅ‘ไธŠ็ฟป้กตๅ‘ไธ‹็ฟป้กต>>
©2008-็Žฐๅœจ cucdc.com ้ซ˜็ญ‰ๆ•™่‚ฒ่ต„่ฎฏ็ฝ‘ ็‰ˆๆƒๆ‰€ๆœ‰