以向量修正法改進傳統六參數法－－附面積約制後之坐標轉換成果

Compare the Improved Results by using Coordinate Vector Correction (CVC) method with Traditional AffineTransformation with Area Constraint

本研究實驗區選定台中市北屯區某一圖解區地段，先以傳統六參數轉換以及附加面積限制之六參數進行坐標轉換後，再使用本文所提出之坐標向量修正法進行各界址點坐標修正，並將成果進行比較與分析。參數坐標轉換不論使用何種模式或是多少參數，皆是將整個轉換區域視為在同一種狀況下進行的坐標調整，但是實際上的誤差與理想的改正值卻沒有被「各別」的去討論，坐標向量修正法便是利用「面積」與「點」的方式對於坐標做進一步的修正，補足了整體坐標轉換所產生之盲點。依據實驗區所得出的數據指出，不論是使用傳統的六參數轉換後或是加入面積約制條件的六參數轉換，使用本文提出之坐標向量修正法可將轉換面積與登記面積之RMSE 值降低至少25%和22%，另外在各宗地超出公差的數量也明顯減少57%及42%；在界址點位移方面，大部分的界址點改正值都不超過±6 ㎝，而超出的情況在本文中也舉例分析並且討論，其原因多為面積差值過大或是幾何狀況極差的狀況下才會產生，應屬於原始資料的謬誤。在坐標轉換後使用本文提出之坐標向量修正法則可以極小的量改正界址點的位置，讓更多的轉換後面積與登記面積不符的問題能夠有效的解決。 The experimental area of this research is located at a section of graphic area in Beitun District, Taichung City. By using traditional affine coordinate transformation and affine transformation witharea constrain, then correcting the boundary points by the method of coordinate vectorcorrection(CVC) method which is presented in this study, and analyses the results.No matter what kind of parameter transformations, they all adjust whole area as one condition.But the errors and ideal adjustments have not been discussed individually. Coordinate Vector Correction(CVC) Method uses ‘Area’and ‘Point’to correct each coordinates in order to improve theblind spot of parameter transformation.The obtained results indicate that the vector correction method no matter in traditional affinetransformation or with area constraint which was presented in this study can improve the differentialvalue from transformed area and registered area up to 25% and 22% at least, and the amount which isover-tolerance land parcels was decreased up to 57% and 42%. In the displacement of boundarypoints, most of the correction value was smaller than ±6 cm. The situations which was over ±6 cmwere also discussed in this study, most reasons are the big differential value and the bad shape andsize in geometry. They should belong the mistakes from the original data.This method displaces few amounts from boundary point, but can solve more problems whenthe transformed area is different form the registered area.

The experimental area of this research is located at a section of graphic area in Beitun District,Taichung City. By using traditional affine coordinate transformation and affine transformation witharea constrain, then correcting the boundary points by the method of coordinate vectorcorrection(CVC) method which is presented in this study, and analyses the results.No matter what kind of parameter transformations, they all adjust whole area as one condition.But the errors and ideal adjustments have not been discussed individually. Coordinate Vector Correction(CVC) Method uses ‘Area’and ‘Point’to correct each coordinates in order to improve theblind spot of parameter transformation.The obtained results indicate that the vector correction method no matter in traditional affinetransformation or with area constraint which was presented in this study can improve the differentialvalue from transformed area and registered area up to 25% and 22% at least, and the amount which isover-tolerance land parcels was decreased up to 57% and 42%. In the displacement of boundarypoints, most of the correction value was smaller than ±6 cm. The situations which was over ±6 cmwere also discussed in this study, most reasons are the big differential value and the bad shape andsize in geometry. They should belong the mistakes from the original data.This method displaces few amounts from boundary point, but can solve more problems whenthe transformed area is different form the registered area.