% * This code was used in the following articles: % * [1] Learning 3-D Scene Structure from a Single Still Image, % * Ashutosh Saxena, Min Sun, Andrew Y. Ng, % * In ICCV workshop on 3D Representation for Recognition (3dRR-07), 2007. % * (best paper) % * [2] 3-D Reconstruction from Sparse Views using Monocular Vision, % * Ashutosh Saxena, Min Sun, Andrew Y. Ng, % * In ICCV workshop on Virtual Representations and Modeling % * of Large-scale environments (VRML), 2007. % * [3] 3-D Depth Reconstruction from a Single Still Image, % * Ashutosh Saxena, Sung H. Chung, Andrew Y. Ng. % * International Journal of Computer Vision (IJCV), Aug 2007. % * [6] Learning Depth from Single Monocular Images, % * Ashutosh Saxena, Sung H. Chung, Andrew Y. Ng. % * In Neural Information Processing Systems (NIPS) 18, 2005. % * % * These articles are available at: % * http://make3d.stanford.edu/publications % * % * We request that you cite the papers [1], [3] and [6] in any of % * your reports that uses this code. % * Further, if you use the code in image3dstiching/ (multiple image version), % * then please cite [2]. % * % * If you use the code in third_party/, then PLEASE CITE and follow the % * LICENSE OF THE CORRESPONDING THIRD PARTY CODE. % * % * Finally, this code is for non-commercial use only. For further % * information and to obtain a copy of the license, see % * % * http://make3d.stanford.edu/publications/code % * % * Also, the software distributed under the License is distributed on an % * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either % * express or implied. See the License for the specific language governing % * permissions and limitations under the License. % * % */ function [NewBox] = RefineBox(Sup, Box, InScribeRatioThre); NumBox = size(Box,1); VOri = Box(:,3) - Box(:,1) + 1; HOri = Box(:,4) - Box(:,2) + 1; AreaOri = VOri.*HOri; NewBox = [zeros(size(Box,1),4)]; OriShapeFlag = 0; ShiftFlag = 1; NumScale = 100; for i = 1:NumBox BoxMask = zeros(size(Sup)); BoxMask(Box(i,1):Box(i,3), Box(i,2):Box(i,4)) = 1; BoxMask = logical( BoxMask); ElementInscribe = setdiff( sort( unique( Sup( BoxMask)) ), 0); ElementOutSide = setdiff( sort( unique( Sup( ~BoxMask)) ), 0); ElementInscribeHistC = sparse(1, max(Sup(:))); ElementOutSideHistC = ElementInscribeHistC; ElementInscribeHistC( ElementInscribe) = histc(Sup( BoxMask),ElementInscribe); ElementOutSideHistC( ElementOutSide) = histc(Sup( ~BoxMask),ElementOutSide); InscribeRatio = ElementInscribeHistC./( ElementInscribeHistC+ElementOutSideHistC); InscribeRatio( isnan( InscribeRatio)) = 0; SupIndRefined = find( InscribeRatio > InScribeRatioThre); SupShape = zeros(size( Sup)); for j = SupIndRefined SupShape( Sup == j) = InscribeRatio(j); end % decide new Box size SupShape(~BoxMask) = 0; Area = sum( SupShape(:) ~=0) Ratio = Area./AreaOri(i); % pick the Shape of the newBox if ~OriShapeFlag Vmax = VOri(i,1); Hmax = HOri(i,1); Vmin = Area/Hmax; count = 1; for Vtrial = linspace(Vmin,Vmax,NumScale) V(count) = max( round( Vtrial), 1); H(count) = max( round( Area/Vtrial), 1); if ShiftFlag h = ones(V(count),H(count)); res = filter2(h,SupShape(Box(i,1):Box(i,3), Box(i,2):Box(i,4)), 'valid'); [CCol ICol ] = max(res, [], 2); [C IRow ] = max(CCol); Shift(count,:) = [IRow ICol(IRow)] - 1; Score(count) = C; else Shift(count,:) = [VOri(i) HOri(i)] - [V(count) H(count)]; Shift(count,:) = round(Shift(count,:)/2) - 1; Score(count) = sum( sum( SupShape( (Box(i,1):(Box(i,1)+V(count)))+Shift(count,1), (Box(i,2):(Box(i,2)+H(count)))+Shift(count,2)))); end count = count + 1; end [BestScore BestIndex] = max(Score); NewBox(i,:) = [Box(i,1)+Shift(BestIndex,1) Box(i,2)+Shift(BestIndex,2) Box(i,1)+Shift(BestIndex,1)+V(BestIndex) Box(i,2)+Shift(BestIndex,2)+H(BestIndex)]; else V = max( round( VOri(i)*Ratio), 1); H = max( round( HOri(i)*Ratio), 1); Shift = [VOri(i) HOri(i)] - [V H]; Shift = round(Shift/2) - 1; NewBox(i,:) = Box(i,:) + [ Shift -Shift]; end end