{"id":439,"date":"2012-01-13T15:50:33","date_gmt":"2012-01-13T20:50:33","guid":{"rendered":"https:\/\/blogs.mathworks.com\/steve\/?p=439"},"modified":"2012-01-13T15:50:33","modified_gmt":"2012-01-13T20:50:33","slug":"five-years-ago-august-september-and-october-2006","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/steve\/2012\/01\/13\/five-years-ago-august-september-and-october-2006\/","title":{"rendered":"Five years ago: August, September, and October 2006"},"content":{"rendered":"
\r\n

Much of the information I posted in this blog years ago is still useful today. Image processing theory hasn't been completely\r\n been overturned since then, and I'm still talking about MATLAB after all. For the benefit of readers who have joined the party\r\n more recently, I will occasionally recap the useful bits that I posted about five years ago.\r\n <\/p>\r\n

In August, September, and October of 2006 I finally finished (mostly) a long series of posts covering spatial transforms<\/a>. I posed a reader challenge<\/a> to come up with a particularly interesting and creative custom spatial transformation. My favorite submission<\/a> was this one:\r\n <\/p>\r\n

<\/p>\r\n

I told the story of how we guessed wrong<\/a> in the early 1990s about which of the several variants of the discrete cosine transform we should use in the Image Processing\r\n Toolbox.\r\n <\/p>\r\n

I poked fun<\/a> at advertising techniques for digital cameras.\r\n <\/p>\r\n

<\/p>\r\n

I taught about several image processing concepts, including Hough transforms<\/a>, separable convolution<\/a>, and nonflat grayscale dilation and erosion<\/a>.\r\n <\/p>\r\n

And I finally ended ten years of public silence about the MATLAB default image<\/a>.\r\n <\/p>\r\n

<\/p>\r\n

<\/p>\r\n

<\/p>\r\n

<\/p>\r\n

<\/p>\r\n

<\/p>\r\n

Be sure to check out the blog archives<\/a> for more oldies but goodies.\r\n <\/p>