When I work on video processing problems one of the tasks I perform most frequently is estimating the velocity of individual pixels\u00a0in a video stream. It was while solving one of these problems that I stumbled\u00a0upon Stanley Chan's submission on sub-pixel motion estimation without interpolation.\r\n
Contents<\/strong><\/h2>\r\n\r\n\r\n\t- Read Input Video<\/a><\/li>\r\n\t
- Setup Motion Estimation Parameter<\/a><\/li>\r\n\t
- Estimate Motion Between First Two Frames<\/a><\/li>\r\n\t
- Quiver plot to view motion vectors<\/a><\/li>\r\n\t
- Other motion estimation techniques<\/a><\/li>\r\n<\/ul>\r\n<\/div>\r\n
Read Input Video<\/strong><\/a><\/h2>\r\nLet's start by reading in a video which we'll use to estimate the motion between individual pixels from frame to frame.\r\nvidReader = VideoReader('visiontraffic.avi'<\/span>,'CurrentTime'<\/span>,11);<\/pre>\r\nSetup Motion Estimation Parameters<\/strong><\/a>\r\n
We then setup the parameters for the motion estimation, in this case we set the block size to 8 and the pixel search limit to 10.<\/h2>\r\nopts.BlockSize = 8;\r\nopts.SearchLimit = 10;\r\n<\/pre>\r\nEstimate Motion Between First Two Frames<\/strong><\/a><\/h2>\r\nNow lets read the first two frames of video and and estimate the motion using the Bidirectional_ME function from the FileExchange entry.\r\nframeRGBLast = readFrame(vidReader);\r\nframeGrayLast = rgb2gray(frameRGBLast);\r\nframeRGB = readFrame(vidReader);\r\nframeGray = rgb2gray(frameRGB);\r\n[MVx, MVy] = Bidirectional_ME(im2double(frameGray), im2double(frameGrayLast)...<\/span>\r\n , opts);\r\n\r\nfigure;\r\nsubplot(2,2,1)\r\nimagesc(MVx.^2);\r\ntitle('Motion Field X-Direction'<\/span>);\r\naxis image<\/span>;\r\nsubplot(2,2,2)\r\nimagesc(MVy.^2);\r\ntitle('Motion Field Y-Direction'<\/span>);\r\naxis image<\/span>;\r\nsubplot(2,2,3);\r\nimshow(frameRGBLast);\r\ntitle('Previous Frame'<\/span>);\r\nsubplot(2,2,4);\r\nimshow(frameRGB);\r\ntitle('Current Frame'<\/span>);\r\n<\/pre>\r\n![\"POTW_Post_01\"](\"https:\/\/blogs.mathworks.com\/pick\/files\/POTW_Post_01.png\")
<\/a>\r\n<\/h2>\r\nQuiver plot to view motion vectors<\/strong><\/a><\/h2>\r\nYou can also view the motion vectors using a quiver plot.\r\nfigure;\r\nquiver(MVx(end:-1:1,:), MVy(end:-1:1,:));\r\naxis image<\/span>;\r\ntitle('Motion Vector Field'<\/span>);\r\n![\"POTW_Post_02\"](\"https:\/\/blogs.mathworks.com\/pick\/files\/POTW_Post_02.png\")
<\/a><\/pre>\r\nOther motion estimation techniques<\/strong><\/a><\/h2>\r\nThe Computer Vision System Toolbox provides several other ways to do motion estimation include some variants of optical flow and block matching. To learn more type\r\ndoc vision<\/span>\r\n<\/pre>\r\n<\/h2>\r\n<\/h2>","protected":false},"excerpt":{"rendered":"<\/div>Avi's pick of the week is Subpixel Motion Estimation without Interpolation by Stanley Chan.\r\n\r\n When I work on video processing problems one of the tasks I perform most frequently is estimating the... read more >><\/a><\/p>","protected":false},"author":132,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[16],"tags":[],"_links":{"self":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807"}],"collection":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/users\/132"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/comments?post=6807"}],"version-history":[{"count":37,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807\/revisions"}],"predecessor-version":[{"id":8214,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807\/revisions\/8214"}],"wp:attachment":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/media?parent=6807"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/categories?post=6807"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/tags?post=6807"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
- \r\n\t
- Read Input Video<\/a><\/li>\r\n\t
- Setup Motion Estimation Parameter<\/a><\/li>\r\n\t
- Estimate Motion Between First Two Frames<\/a><\/li>\r\n\t
- Quiver plot to view motion vectors<\/a><\/li>\r\n\t
- Other motion estimation techniques<\/a><\/li>\r\n<\/ul>\r\n<\/div>\r\n
Read Input Video<\/strong><\/a><\/h2>\r\nLet's start by reading in a video which we'll use to estimate the motion between individual pixels from frame to frame.\r\n
vidReader = VideoReader('visiontraffic.avi'<\/span>,'CurrentTime'<\/span>,11);<\/pre>\r\nSetup Motion Estimation Parameters<\/strong><\/a>\r\n
We then setup the parameters for the motion estimation, in this case we set the block size to 8 and the pixel search limit to 10.<\/h2>\r\nopts.BlockSize = 8;\r\nopts.SearchLimit = 10;\r\n<\/pre>\r\n
Estimate Motion Between First Two Frames<\/strong><\/a><\/h2>\r\nNow lets read the first two frames of video and and estimate the motion using the Bidirectional_ME function from the FileExchange entry.\r\n
frameRGBLast = readFrame(vidReader);\r\nframeGrayLast = rgb2gray(frameRGBLast);\r\nframeRGB = readFrame(vidReader);\r\nframeGray = rgb2gray(frameRGB);\r\n[MVx, MVy] = Bidirectional_ME(im2double(frameGray), im2double(frameGrayLast)...<\/span>\r\n , opts);\r\n\r\nfigure;\r\nsubplot(2,2,1)\r\nimagesc(MVx.^2);\r\ntitle('Motion Field X-Direction'<\/span>);\r\naxis image<\/span>;\r\nsubplot(2,2,2)\r\nimagesc(MVy.^2);\r\ntitle('Motion Field Y-Direction'<\/span>);\r\naxis image<\/span>;\r\nsubplot(2,2,3);\r\nimshow(frameRGBLast);\r\ntitle('Previous Frame'<\/span>);\r\nsubplot(2,2,4);\r\nimshow(frameRGB);\r\ntitle('Current Frame'<\/span>);\r\n<\/pre>\r\n<\/a>\r\n<\/h2>\r\n
Quiver plot to view motion vectors<\/strong><\/a><\/h2>\r\nYou can also view the motion vectors using a quiver plot.\r\n
figure;\r\nquiver(MVx(end:-1:1,:), MVy(end:-1:1,:));\r\naxis image<\/span>;\r\ntitle('Motion Vector Field'<\/span>);\r\n<\/a><\/pre>\r\nOther motion estimation techniques<\/strong><\/a><\/h2>\r\nThe Computer Vision System Toolbox provides several other ways to do motion estimation include some variants of optical flow and block matching. To learn more type\r\n
doc vision<\/span>\r\n<\/pre>\r\n<\/h2>\r\n
<\/h2>","protected":false},"excerpt":{"rendered":"
<\/div>Avi's pick of the week is Subpixel Motion Estimation without Interpolation by Stanley Chan.\r\n\r\n When I work on video processing problems one of the tasks I perform most frequently is estimating the... read more >><\/a><\/p>","protected":false},"author":132,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[16],"tags":[],"_links":{"self":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807"}],"collection":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/users\/132"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/comments?post=6807"}],"version-history":[{"count":37,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807\/revisions"}],"predecessor-version":[{"id":8214,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/posts\/6807\/revisions\/8214"}],"wp:attachment":[{"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/media?parent=6807"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/categories?post=6807"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/pick\/wp-json\/wp\/v2\/tags?post=6807"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
- Setup Motion Estimation Parameter<\/a><\/li>\r\n\t