Electronic Photography on the Microscope
With an electronic camera and a microscope, it would seem that all the parts necessary for electronic photography are present. Actually, some type of controlling device and storage device for the images is necessary. This generally takes the form of a desktop personal computer. The presence of a computer system provides the necessary means to control the various camera functions, and also provides a medium on which to store images. Assuming adequate space on the computers storage disc, almost any computer can be used for camera control and image storage. Particular considerations are the availability of a framegrabber that will fit in the computer, if the particular camera uses a framegrabber, or a digital interface card that fits in the computer if the camera has direct digital input. With the continued development of high speed serial busses such as Firewire, and support for this bus structure in the latest version of computer operating systems, both PC and Mac, it may soon become even easier to connect digital cameras to computers. In addition to controlling the camera, the computer also generally provides a display of the captured image. Most camera software will include an image display function, so the user can verify that the image obtained was what was desired from the specimen. If the image is satisfactory, then the software will permit the user to store the image, usually with a name chosen by the user. Images are stored using files of type bitmap. Bitmap files are made up of individual blocks, called pixels, or picture elements. Each pixel represents a particular portion of the image. In general, at standard magnification of the display image, individual pixels will not be seen. If the display software permits zooming a common feature, the displayed image can be enlarged until individual pixels are seen.
These will be recognized because the image will take on a mosaic or stair step appearance. There are a number of specific formats that can be used for bitmap type image files. In the Windows operating system, there is a file type simply called bitmap with the file extension .BMP. For scientific images, a more common bitmap file format is Tagged Image Format, or .TIF. There are many other types of bitmap files, such as .GIF, which is commonly used to display images on the Internet, and a newer replacement for GIF called .PNG. From the standpoint of the user, these various file formats are essentially equivalent, since all can store an image and display an image, assuming software that understands each flavor of bitmap. The major differences between these various image file types is the type and arrangement of information in the file “header” which is a part of the file that tells any software that displays it specific information about the image contained in the file. Typically the header will have information such as the size of the image in x and y directions, the amount of information in each pixel, whether the image is monochrome or color, and other pertinent information, depending on the particular file format. There is another category of image file formats called vector images.
These images are not produced by cameras, but by graphic arts type programs. In a vector file, each element is mathematically defined. As an example, in a vector file, a line is simply defined as the x and y coordinates of the start of the line, and the x and y coordinates of the end of the line. In a displayed vector file, a line will be very straight, just like one drawn on paper. In contrast, a straight line that runs diagonally across the screen in a bitmap file will have a stair step appearance. Since cameras do not produce vector files although some graphics programs can convert bitmap files to vector images we will not discuss vector files any further. With such large images, it is only natural to consider ways to make the image size smaller. This would be convenient in order to fit more images in a given space, to move images from one computer to another, and to speed up the storage and retrieval of images. It is possible to compress images. There are many ways to compress image. Compression techniques fall into two distinct categories: lossy and lossless. As is apparent from the name, a lossy technique actually throws away part of the image information. Once this happens, it is gone forever, and cannot be retrieved. In lossless compression, only information that can be reconstructed is tossed out. With lossless compression, the original image can be restored. An example of lossless compression is a type of compression called run length encoding. In this technique, each pixel is examined in turn.
As soon as three or more adjacent pixels are seen to have the same value, the string of identical pixels are replaced by two numbers, the value of each of the pixels, followed by the total number of adjacent pixels with that value. It should be obvious that run length encoding could make images files much smaller if the image contains large areas of identical pixel values that are adjacent to each other. An image that has large solid areas such a bone spicules might fit this description or an image with large open spaces. Unfortunately, most cell and tissue specimens do not have large areas of identical pixels, and therefore run length encoding generally does not provide much saving of storage space for these type of images. As a general rule, compression of images of microscopic objects should be avoided unless the compression is lossless. The cost of additional storage is a small price to pay for retention of all image information. After all, we never know what new analysis techniques will permit us to extract from previously obtained images.
