# ES 555 Lab Exercise on Digital Image Scanning

## Introduction

Scanning of photographs is based on two kinds of input imagery—original film (negative or positive) and photographic prints. For SFAP, the former normally means 35-mm (or 70-mm) film, and the latter are usually prints of 4x6-inch (10x15-cm) size or larger. Begin by examining the nominal dimensions of a photograph on original 35-mm film.

 The film is 35 mm wide, including the margins with sprocket holes. The image is 24 mm wide (between sprocket holes) and 36 mm long. The image, thus, has a length-to-width ratio of 3:2. This example is color-slide film; the same dimensions apply to 35-mm color-print film. Kite aerial photograph near Halstead, KS, © J.S. Aber. This view shows the same image as mounted in a standard color-slide frame (black). The frame "window" measures 23 by 35 mm, thus, 1 mm is obscured in both width and length. In most cases, this loss of marginal data is trivial. Kite aerial photograph near Halstead, KS, © J.S. Aber.

## Exercise

Scanning resolution is usually expressed in terms of dots per inch (dpi) or dots per centimeter. There are 2.54 cm (or 25.4 mm) per inch.

1. What is the nominal size of a photograph in inches (on 35-mm film)? Round your answer to 1/100th of an inch.

2. Now assume the film picture will be scanned at 400 dpi. How many rows (width) and columns (length) would result in the image? How many total pixels does this represent?

The simplist format for saving a digital image file is called bitmap (bmp), in which a color value is saved for every pixel in the image. In the case of the previous example, this would result in a file more than 200 kb in size. This may be the most accurate means for digital image storage, but for bigger images it results in excessively large files. Most image files are saved using some kind of compression algorithm, which reduces file size. This is possible because many images contain much redundant information—adjacent pixels with the same value. There are two image compression formats most common in useage (Weinman and Heavin 1997).

• Graphic Interchange Format (GIF): An 8-bit (28) format which can contain only 256 colors. GIF compression is "lossless" assuming the original image has no more than 256 colors. Versions of GIF support transparency, interlacing, and animation. GIF is best for compressing graphic illustrations.

• Joint Photographic Experts Group (JPEG): A 24-bit (3 x 28) format which can contain millions of colors. JPEG was developed specifically for photographs. It is a "lossy" compression that removes information from the image file; however, this data loss is only slightly visible in most cases.

 Color slide scanned at 400 dpi and saved in jpeg format (high quality). Click on the small image to see full-sized version (32 kb). Kite aerial photograph near Halstead, KS, © J.S. Aber. Color slide scanned at 600 dpi and saved in jpeg format (high quality). Click on the small image to see full-sized version (59 kb). Kite aerial photograph near Halstead, KS, © J.S. Aber.

Compare the two images above, scanned at different dpi from the same original mounted color-slide, and saved with the same format. Note the differences in display size and detail of features depicted in the two images.

3. When scanning dpi is increased by 50% (from 400 to 600, for example), what happens to the number of rows and columns and the total number of pixels in the image? What is the result in terms of file size?

For high-resolution digital images, film negatives and slides are often scanned at resolutions as high as 2400 dpi. This tiny scanning resolution approaches the size of silver halide crystals in the film emulsion, and represents a practical resolution limit for conventional film.

4. For scanning film at 2400 dpi, what is the actual size (linear dimension) for a cell (pixel) in the scanned image? Give your answer in inches and mm.

Now turn your attention to photographic prints, which are normally enlarged several times in comparison to the original film. A typical print size (for 35-mm film) is approximately 4 by 6 inches (10 by 15 cm), which preserves the length-to-width ratio of 3:2.

5. How much is a 4x6 print enlarged in comparison to the original color film? Give your answer in terms of linear dimension (width and length), not in area.

6. What dpi should be utilized in scanning a 4x6 print in order to achieve results similar to scanning the original film at 600 dpi?

## Scanning procedure

Each student will select one original film (positive or negative) picture or one photographic print for actual scanning. On-campus students may utilize scanning equipment in the GeoSpatial Analysis laboratory. Distance-learning students will have to locate appropriate scanning equipment on their own. Create a scanned image from the original photograph (film or print). Any subject matter is acceptable—it does not have to be an aerial photograph, and students may work with either color or b/w photography.

Experiment with different factors for scanning and saving images, including various scan resolutions (dpi) and compression formats. In addition, students are encouraged to utilize some simple image processing techniques, which are available with Adobe Photoshop or other image-processing software. Such techniques include the following.

• Cropping and rotating images.
• Adjusting brightness level and color balance.
• Sizing images—enlarging and reducing.
• Smudging or blurring (to remove dust specks, etc.).
• Sharpening and/or edge enhancement.

You should ultimately produce an image that you feel is a good digital version of the original photograph; perhaps even an improvement compared to the original film or print image. Describe briefly the techniques you employed to create the digital image and submit it with your answers.

### Turn in

• Written answers (1-6) and description of techniques.
• Scanned image in digital format.

### Reference

• Weinman, L. and Heavin, B. 1997. <coloring web graphics.2>. New Riders Publ., Indianapolis, 314 p.