Hình 9-2. Dữ liệu từ hình 9-1 trình bày về sự khác biệt trung bình: một đối tượng (một bàn tay) đi qua mặt trước của máy ảnh hơi tối hơn, và độ sáng của đối tượng đó được thể hiện trong đồ thị Phương pháp trung bình sử dụng của bốn OpenCV thói quen: cvAcc (), để tích lũy | Figure 12-15. A fixed disparity forms a plane affixed distance from the cameras some features found on the left cannot be found on the right but the ordering of those features that are found remains the same. Similarly there may be many features on the right that were not identified on the left these are called insertions but insertions do not change th e order offoatuces although t hey may spread tho se features out. The procedure illustrated in Figure 12-16 reflects the ordering constraint when matching features onahor ntali can fin e. Given the rmrllesn cMnweadiaoarity iacrementAd wecaaretermira smrUcst anniev-aier devihrrnga natol-ifion nz ny vtiit-g- the formula -e AZ nsn-bi fT Itis usafultakeeu this formnloin minn sothat yoy kruw ialnil tied nfdeath resolution io .rxp ct fromyour ntcaen rỏ. After correspondence we turn to postfiltering. The lower part otHgure 12-13 shows a typical matching function response as a feature is swept from the minimum disparity out to maximum disparity. Note that matches often have the characteristic of a strong cent cal peak turrottndc dba oide totes-Onc a rot enoe aanr Writ featn re corraviionnlcncnn ihntwcrn thn twe ofow-1 poclfiltering is used to prevent false matches. OpenCV makes use bf ths mated-ig -urrttob pattern via c nniquerensRatCo parameter twhose VnSrult veiun it toa thntPHnns nnt -oetehe-i whsrr cs-qr enrr-Catsc a maach_valeain_octchae min match. 44- I niapter1s P sr nrmlW Vteisi Figure 12-16. stereo correspondence starts by assigning point matches between corresponding rows in the left and right images left and right images of a lamp upper panel an enlargement of a single scan line middle panel visualization of the correspondences assigned lower panel . To make sure that there is enough texture to overcome random noise during matching OpenCV also employs a textureThreshold. This is just a limit on the SAD window response such tiiat no match is considered whose response is below the textereTSreehold the defadhevslue
