You are here: Technology Image processing Put in a word for the luckless scanner…

Put in a word for the luckless scanner…

Rate this item
(1 Vote)

Grotty results of hard-substratum map scanning with a $8.000 scanner set us thinking about use of inexpensive office equipment.

We had to scan a lot of maps on the hard substratum in the course of a project. About one half of them were on the usual aluminum foil but the rest were “monsters” up to 1.5 m long on the fibreboard substratum. Metal sheets often had turned-up angles and their shape resembled something like carriage springs but that was nothing in comparison with exotic forms of maps on the fibreboard bottom-layer. Besides, in my long practice I had never seen a flatbed scanner with the glass big enough for such giants...

Even now I shudder to recall our tries to scan these maps with a large-format roller scanner. Imagine a heavy sheet bent under gravity that is slowly creeping into the feeder slit. Two strong guys carefully support it from both sides of the scanner while the pressure rollers try to straighten the “spring” and demonstrate clutch slip at any opportunity. These “high tech exercises” generated fatal scanning errors up to 1.5 centimeters (!!!) .

But we had already signed the contract and could not give it up. So, we began to look for a way out.

An office scanner vs. a large-format "construction"

We fixed upon an inexpensive office scanner Mustek ScanExpress P of A3 format (~$150, Simplicity of the head-moving mechanism caused anxiety first but didn’t become a source of significant inaccuracy – maximal error was less than 0.3 mm when measured on a 20 mm grid (less than 0.1%).

We installed the scanner flush with the table but that was a bad idea – test scanning provided dark and blurred images because of the equipment design features .

Fig.1. It is impossible to get a contrast scan applying a common mirrorless flatbed scanner because of some design features of the equipment.

Most modern inexpensive table-top scanners have a mirrorless optical system with the light-sensitive devise (CCD) placed in immediate proximity to the glass. It makes the entire design cheaper but the lens of CCD has a very short focal distance and thus the zone of contrast imaging is situated just above the glass surface (red dash-and-dot line). At that, the glass of all office scanners is recessed a little into the scanner case. The sheet “hovering” above the glass does not get into the contrast zone and the resulting image is unfit for further work. We decided to modify the case design as it was the only source of the problem.

Fig.2. General view of the scanner after modification.

General view of the modified scanner is represented in Figure 2. Scanner without the top cover (3) is rigidly fastened on a hard basement of laminated fibreboard (6) 720 x 1000 х 18 mm large. The scanner is placed in a corner to facilitate scanning of large sheets. The construction is covered with a glass (1) glued to a rectangular frame (5) also made of fibreboard. Glass width is 2.5 – 3.0 mm (as in the scanner) but it is not necessary a mirror glass – a good windowpane will do. Additional supports (2) glued to the inner surface of the glass to avoid sagging and increase its “bearing capacity” rest upon the basement.

Scanner calibration marks (7) and additional marks for sheet positioning (4) are also glued on the glass. Two diagonally placed metal dowels 5 mm in diameter (missing in the figure) are set fast into the basement. They go into holes in the frame to fix the later relatively the basement.

All given sizes may be changed but there are two points that deserve to be considered in more details.

Frame making

As a rule, the scanning head has a spring-loaded plate to provide good contact with the glass inner surface. It is a necessary condition for contrast scanning in the entire working field. To determine the height of the frame exactly, place the scanner on a flat foundation and measure the distance H from the later to the upper surface of the head in the lower (pressed) position (see Fig. 3). Then measure displacement of the head’s mobile part . Height of the frame walls is . Manufacturing method should guarantee tolerance no more than Non-compliance with this condition may cause head seizing or contrast loss.

Fig.3. Determining the frame height.

We recommend the following assembly sequence that does not require high accuracy at processing of mating parts:

  1. Delete laminar film from junctions of the frame ready parts and clean the surface;
  2. Drill holes for screws in joints as it is shown in Fig. 4;
  3. Fulfill draft assembling of the frame but do not tighten the screws completely. It should be possible to move frame parts at joints a little;
  4. Put the frame on a flat surface of the table and place the future basement of the scanner and a weight atop it (Fig. 4);
  5. Adjust angles of the frame, check the distance between the upper plate and the table. Refine construction parts if necessary;
  6. Dismount the weight and the basement, loose the screws, and fill gaps with epoxy adhesive;
  7. Tighten the screws but keep frame parts mobile relatively each other;
  8. Put the frame on the table and press it again with the basement and the weight;
  9. Repeatedly adjust angles of the frame and check the distance between the upper plate and the table. Delete the glue extruded from joints with an acetone-wetted wad;
  10. The frame is ready after 16-18 hours required for hardening of epoxy adhesive at room temperature.
Fig. 4. Frame assembling. Glue the glass down in the same way.

Glue the glass down in the same way. Use a flint paper to depolish a band about 2 cm wide around the glass periphery for glue laying.  Depolish also the zones where additional supports will be fastened. Thoroughly clean the glass with ethanol and dry it out. Glue protective paper on the face side with rubber-based adhesive smeared around the glass periphery. Put the glass on the table protected face down and ungrease depolished zones with acetone.

Attention! Depolish the glass wet (add some water) applying water-resistant flint paper to protect your eyes and respiratory apparatus from dangerous dust.

Put a rather thick (several mm) layer of epoxy adhesive on end-face surfaces of the frame, wait for 5-10 minutes, turn the frame (adhesive layer down), and put it on the glass. Wait for 10 minutes again, put the basement and the weight atop the frame, and delete surplus of adhesive.

Fig.5. Placing of additional supports.

Figure 5 represents the final stage of assembling that consists in placing of additional supports. Make these supports 1-2 mm shorter than the frame walls as they should not touch the glass at frame installation on the basement. To glue the supports, act as follows:

  1. Mark the sites of support attaching on the basement.
  2. Put a drop of rubber-based adhesive on one face plane of every support and place the later on the basement.
  3. After all supports are fastened, apply a 3-5 mm thick layer of epoxy adhesive on their upper face planes.
  4. Use acetone to ungrease depolished zones of the glass where additional supports will be attached to.
  5. Cautiously place the frame with the glass on the basement.

Attention! Positional relationship of the frame and the basement should remain unchanged after fastening of the additional supports

Calibration marks

These marks are necessary at device provisioning for determination of the working field, initial positioning of the scanning head, and setting of white and black color levels. As a rule, these is a mark parallel to the glass short side, placed between the glass and the plastic case at the site of head parking. Exact reproducing of this mark is very likely the most critical part of the whole venture.

Fig.6. Calibration marks of the scanner Mustek ScanExpress P A3.

Marks of different scanners may differ; ones of Mustek ScanExpress A3 are shown in the figure. Some models have also a mark parallel to the long side of the glass. It is important to keep mark size and position relatively the scanning mechanism at copying. Incorrect positioning of calibration marks may cause devise malfunction and badly decrease contrast of scans.

The mark should be put on the correct surface of the glass relatively the scanning head (as it was in the scanner) to avoid brightness and color distortions. Non-compliance with this requirement results in dark and greenish images.

To make the marks, we used pressure sensitive adhesive Oracal 641 010M and 070M ( Most likely, you will have to fasten it on the glass glue layer down (to the scanner), so make sure that there are no dust particles and air bubbles between the mark and the glass – they may cause vertical bands in scans.

Fig.7. Fastening the pressure sensitive film.

Glue a bigger that necessary piece of white film first and then draw contours of the mark with a B-pencil on it. Figure 7 illustrates how to avoid dirt and air bubbles under the film: turn up the protective layer and fasten a narrow band along the short side of the pressure sensitive film; then hold the rest piece with one hand and smooth the fasten part with the other as it is indicated by the arrows.

Now gash the film with a sharp knife along the drawn contours and delete the cores. Clear away remnants of the glue layer with a wad wetted in white spirit, then delete white spirit with a detergent. Cut out insets from black film and paste them into the prepared sites. It may be reasonable also to glue additional narrow marks indicating the field of scanning – they will help to place map sheets correctly (Fig. 2).

Map scanning

Scan the map in several intersecting fragments (zones) as shown in Figure 8. There should be at least four grid nodes (ticks) in every fragment. It is convenient to mark the zones with a pencil on the back of every sheet and to number them. It helps to place the sheet correctly and to avoid excess scanning.

Fig.8. Fragmentation of a large map sheet. Ticks are marked with crosses.

See to good contact of the map with the scanner glass and use additional weights if necessary. We loaded it with steel washers, up to 25 kg in sum..

Adding of resultant images to the project is described in Chapter 5 of the Easy Trace User Guide. It is possible to crop fragments along their common ticks together with correction of geometrical distortions. For example, Fragment 1 may be cropped on the right and below, Fragment 2 – on the right, on the left and below, and so on.


Average efficiency of a working place equipped with this scanner is about 10-15 sheets 50 x 50 cm (4-6 sq. m) per eight-hour working day, including zone marking on the sheets, scanning, and adding of resultant images to the project.

More in this category: Next »

Add comment

Security code