Try checking the range of your image data. 'imshow' and 'implay' will use a default display range for the data based on the data type. For example, if 'img' is an single or double type, then the display range is [0,1], where 0 maps to black and 1 maps to white. If 'img' is an integer type, the default display range will be the minimum and maximum representable values for that integer class. As an example, a uint16 array would have the range [0,65535].
If your data falls outside of the default range, you need to convert it. If you are using 'imshow', you can do so by specifying the 'DisplayRange' parameter. However, 'implay' does not have a parameter to adjust the display range so you will need to convert the matrix to a grayscale image. You can do so using 'mat2gray' and specifying the 'amin' and 'amax' parameters.