{"id":122,"date":"2007-03-16T07:00:59","date_gmt":"2007-03-16T11:00:59","guid":{"rendered":"https:\/\/blogs.mathworks.com\/steve\/2007\/03\/16\/computing-distance-using-image-file-information\/"},"modified":"2019-10-23T08:46:20","modified_gmt":"2019-10-23T12:46:20","slug":"computing-distance-using-image-file-information","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/steve\/2007\/03\/16\/computing-distance-using-image-file-information\/","title":{"rendered":"Computing distance using image file information"},"content":{"rendered":"
\n

Patrick asked<\/a> for an example on how to compute distances between objects based on resolution information stored in a file. The TIFF format
\nis one that can store resolution information, and the imfinfo<\/tt> function can tell you about it.<\/p>\n

Let's look at a simple example:<\/p>\n

url = 'https:\/\/blogs.mathworks.com\/images\/steve\/122\/blobs.tif'<\/span>;\r\nbw = imread(url);\r\nimshow(bw)<\/pre>\n
 <\/p>\n
Here's the output of imfinfo<\/tt>:<\/p>\n
info = imfinfo(url)<\/pre>\n
info = \r\n\r\n                     Filename: 'C:\\TEMP\\tp281629'\r\n                  FileModDate: '06-Mar-2007 09:28:16'\r\n                     FileSize: 2176\r\n                       Format: 'tif'\r\n                FormatVersion: []\r\n                        Width: 308\r\n                       Height: 242\r\n                     BitDepth: 1\r\n                    ColorType: 'grayscale'\r\n              FormatSignature: [73 73 42 0]\r\n                    ByteOrder: 'little-endian'\r\n               NewSubFileType: 0\r\n                BitsPerSample: 1\r\n                  Compression: 'CCITT 1D'\r\n    PhotometricInterpretation: 'WhiteIsZero'\r\n                 StripOffsets: [10x1 double]\r\n              SamplesPerPixel: 1\r\n                 RowsPerStrip: 26\r\n              StripByteCounts: [10x1 double]\r\n                  XResolution: 100\r\n                  YResolution: 100\r\n               ResolutionUnit: 'Inch'\r\n                     Colormap: []\r\n          PlanarConfiguration: 'Chunky'\r\n                    TileWidth: []\r\n                   TileLength: []\r\n                  TileOffsets: []\r\n               TileByteCounts: []\r\n                  Orientation: 1\r\n                    FillOrder: 1\r\n             GrayResponseUnit: 0.0100\r\n               MaxSampleValue: 1\r\n               MinSampleValue: 0\r\n                 Thresholding: 1\r\n\r\n<\/pre>\n
Notice the XResolution and YResolution fields, as well as the ResolutionUnit field. According to the TIFF specification, the XResolution and YResolution fields are the number of pixels per resolution unit.<\/p>\n
info.XResolution<\/pre>\n
ans =\r\n\r\n   100\r\n\r\n<\/pre>\n
info.YResolution<\/pre>\n
ans =\r\n\r\n   100\r\n\r\n<\/pre>\n
info.ResolutionUnit<\/pre>\n
ans =\r\n\r\nInch\r\n\r\n<\/pre>\n
Now let's use bwlabel<\/tt> and regionprops<\/tt> to compute the centroids of the objects.<\/p>\n
L = bwlabel(bw);\r\ns = regionprops(L, 'Centroid'<\/span>)<\/pre>\n
s = \r\n\r\n4x1 struct array with fields:\r\n    Centroid\r\n\r\n<\/pre>\n
Compute the distance (in pixel units) between the first two objects.<\/p>\n
delta_x = s(1).Centroid(1) - s(2).Centroid(1);\r\ndelta_y = s(1).Centroid(2) - s(2).Centroid(2);\r\npixel_distance = hypot(delta_x, delta_y)<\/pre>\n
pixel_distance =\r\n\r\n  103.6742\r\n\r\n<\/pre>\n
Finally, use the resolution information from the file to convert to physical distance. (Note: this calculation assumes that
\nthe horizontal and vertical resolutions are the same.)<\/p>\n
physical_distance = pixel_distance \/ info.XResolution<\/pre>\n
physical_distance =\r\n\r\n    1.0367\r\n\r\n<\/pre>\n