Image content impacts how you sharpen images. There are essentially three types of image content: low-frequency, high frequency and a mix of both. (Well that was interesting. This got published in error, my error of course.) Oh well, moving on.

low-frequency image

The above image is all about smoothness and yet still has a graphical quality. This makes this a low-frequency image.

High-Frequency Image

The image of the tree and rocks is considered high-frequency. It has lots of different edge details and lots of edges requiring sharpening.

Medium Frequency Image

This image is a mix of low-frequency sky and clouds and high-frequency trees and grass. Therefore, this is referred to as a medium frequency image.

You images will fall into one of these categories. The point to take away here is that one stop sharpening does not always work. I know there are many who disagree, but read on and you’ll see what I mean.

Digital image sharpening is done to pixels. By “sharpening” the pixel you create a halo around the pixel to increase or decrease its contrast compared to the pixel next to it. The size of the halo determines the degree of sharpness. And in Lightroom and Photoshop the Radius setting determines the pixel’s halo size.

Therefore, when sharpening a digital image you need to consider a couple of things. Image content impacts the Radius selection. Image capture affects sharpening (image noise, poorly focused lens, chromatic aberration and camera movement). Remember, sharpening cannot repair an out of focus image. Also, you need to consider image enlargement before sharpening.

A high frequency image requires a Radius setting of 0.8 – 0.6. Now these are not absolutes, but guidelines.

A low frequency image requires a Radius setting of 1.5 – 2.0.

The Amount setting affects the intensity of the sharpening.

Given that your computer monitor has a very low resolution, it is really difficult to judge image sharpness by only looking at the monitor. Running a test print is the only true way to judge if the sharpening is adequate or too over the top. Like you, I look at the monitor with an image at 1:1 magnification and “judge” the quality of the image’s sharpness. I then run a print. Sometimes an image looks really great on-screen and the print is too sharp. Sometimes the reverse is also true. So make a test print to evaluate your sharpening.

Also consider setting up some sharpening defaults within Lightroom. It saves a lot of time.

Westworld is a 1973 science fiction / thriller film written and directed by novelist Michael Crichton and starring Yul Brynner, Richard Benjamin, and James Brolin. Set in a high tech amusement park, Peter Martin and John Blane play cowboys in a faux western world with lifelike robots, but when the park’s central computer malfunctions, the vacation turns from harmless fun to potentially fatal, and the friends must figure out how to escape with their lives.

Westworld was the last movie MGM produced before dissolving its releasing company, and was the first theatrical feature directed by Crichton. It was also the first feature film to use digital image processing to pixellate photography to simulate an android point of view. The film was nominated for Hugo, Nebula and Golden Scroll (aka Saturn) awards, and was followed by a sequel film, Futureworld, and a short-lived television series, Beyond Westworld.

Trivia:

• When the Gunslinger robot is splashed in the face with acid, Yul Brynner’s face was coated with an oil-based makeup mixed with ground Alka-Seltzer. A splash of water then produced the fizzing effect.
• The first use of computer digitized images as part of a feature film (not merely monitor graphics) was the Gunslinger’s point of view in Westworld. After the process was finally developed enough to produce satisfactory results, it took a mere eight hours to produce each ten seconds of Gunslinger’s pixellated POV.
• The robot that Yul Brynner portrays is an homage/spoof of his character Chris from The Magnificent Seven (1960) and wears the same costume.
• Michael Crichton became inspired to write this film after a trip to Disneyland, where he saw the Pirates of the Caribbean ride, and was impressed by the animatronic characters.
• Director John Carpenter based the “indestructable” nature of his killer Michael Meyers in Halloween (1978) on Yul Brynner’s character in this film.
• Yul Brynner’s character is known as The Gunslinger. Interestingly, the word “gunslinger” was only created in the 1950s and has no connection.

When shooting in strong light, have you ever noticed the edges of objects seem to have colour shift? This is usually visible at 1:1 enlargement. The most common colour shifts are red or blue.

The most common cause is using the smallest aperture setting on a lens. Occasionally, the maximum aperture value can also cause image quality issues. Ideally, you should perform test shots starting at wide open and then closing down in one stop increments. Then load the files in your image processing software and then enlarge them to 1:1.

Look at the corners of the frames and see is the focus begins to fall off as you approaches the corners. Also check for red/green colour shift at edges of objects. These are two type of lens aberration.

The colour shift can be eliminated within Lightroom.

The loss of focus can be cured by not using the extreme aperture values. You’ll probably find that each lens is the sharpest any where from 1 – 2 stops closed down and about 2 stops from minimum aperture. Again, do test shots to confirm which aperture for each lens gives you the best depth or field and the best edge to edge sharpness.

There are some other things you can do to improve image quality in the camera. If your camera permits Mirror Lock Up (MLU), use it. Buy an electronic cable release and use it too. Always use the lowest ISO possible that gives you the required DoF you want. Buy a good quality tripod. The tripod should weigh about twice as much at the heaviest camera/lens combination. Place a small sand bag on top of the lens of camera body when using a telephoto lens. This will dampen camera shake even more. Use ETTR exposure techniques to maximize image detail in the highlights. Remember the first highlight stop captures 4068 brightness tones. The next stop to the left captures half as many, or 2034 brightness values. The third stop to the left captures 1017 brightness values. So as you can see, the first three stops capture the majority of the image’s brightness values. If you are under exposing or not even using the right hand most zone of the histogram, you are throwing away almost half your image’s most important values.

You might also want to check out RawWorkflow.com. Their Lens Align took, while not cheap, works. I know; I bought it and found out my Canon L series lens was just a bit off on its AF. A simple correction and now the AF feature works really well.

Well there are some things to consider when working on your shooting technique.

Image noise is a real issue with entry-level DLSR cameras. So here is a simple test that will show you how much image noise your camera type/model creates.

Select a typical scene (Just what is a “typical scene”, you may ask. The scene should have at least a 7 – 8 stop contrast range.), set you camera’s auto exposure bracketing (AEB) menu to the following values – normal, +1.33 stops and -1.33 stops and take the shot. (The camera records three images – one at your “normal” exposure, one underexposed by 1.33 stops and one overexposed by 1.33 stops.) You should shoot the images as RAW files.

Down load the images to your computer.  For the underexposed image, increase the exposure slider by about 1.3 – 1.5 stops. For the normal exposure, check the shadows and see if you need to increase them at all. For the overexposed image, reduce the exposure about .5 stops and use the other sliders to bring the shadows back to what you feel is correct. Ideally, all three image’s Histograms should look pretty much the same.

Now enlarge all three images to 1:1 and check out the shadow detail. I’ll bet the underexposed image’s shadow is the most noisy; the normal exposure is somewhat less noisy and the overexposed image’s shadows have the least amount of noise.

This tells you something about image noise and your camera’s sensor. By increasing the exposure (increasing the number of photons that hit the sensor), you increase the signal to noise ratio and thereby decrease the apparent image noise. It also demonstrates the value of the  “exposing to the right” (ETTR) concept. ETTR works because the brightest stop of data, or the right stop of the Histogram contains 2048 levels of brightness data. The lower data stops (the ones to the middle and left) have significantly fewer brightness levels. In fact, as you move each stop to the left, you loose half the data brightness levels.

If your image Histograms consistently DO NOT use the right side of the Histogram then you a loosing image detail and increasing the potential noise in your images.

You need to do some tests of your camera’s sensor to find out just how much ETTR it can handle. Remember, just because the image displayed on your screen indicates highlight clipping (flashing red area) does not mean you have overexposed the shot. When shooting RAW, you can recover the highlights. You need to do a series of .33 stop over exposures until you reach about 1.33 – 2.33 stops overexposure. Then download the RAW files and see which overexposed image you cannot recover the highlight details in by using the Exposure slider, the Recovery slider and the Tone Curve sliders. The image with .33 stops less exposure is the true limit of your sensor, e.g. 1.66. This means you can “safely overexpose” a RAW file image by 1.66 stops knowing you can recover the highlight detail. It also means you have captured the maximum image detail available.

One caveat is that if the sun is shinning as in no clouds then ETTR does not really work. When the sun is out on a cloudless day the contrast range is beyond the range of you sensor to handle all the data detail available in the brightest highlights or the deepest shadows.

While all the above applies to RAW files, what about JPEG files?

Because JPEG files are compressed, even if you apply the in-camera sharpening, so the image noise is more apparent. If you compare a RAW file and a well exposed JPEG file of the same image side by side at normal size, the files may look exactly the same. It is when you begin to enlarge the files that the difference becomes apparent. The JPEG file displays image noise after very little enlargement, where as you can enlarge the RAW file much more before you begin to see any image noise.

On a side note: ETTR means increasing the number of photons hitting the sensor by increasing the exposure. Increasing an exposure means opening up to a larger aperture or using a slower shutter speed. Increasing the ISO does not increase exposure. Whether you shoot an image at ISO 100 or at 400, you still have to zero the meter and so the amount of light (photons) hitting the sensor is the same.

Next, I’ll take a look at Chromatic Aberrations (CA). This is a fancy term for the colour shift you some times see at the edges of highlight objects.

It seems there are three types of sharpening that you can apply to your digital files – capture, creative/selective and output sharpening. Capture sharpening is applied when download a file, or series of files, to your computer. Creative/selective sharpening is applied during the creative manipulation you do to the file to render it the way you want, e.g. crop, adjust exposure, modify tone curves, etc. Output sharpening is what you do when you send a file to print or for screen/web display.

Each of the three sharpening types are handled differently and use different tools in different software applications.

So as you can see, the one stop, or step, sharpening approach may work some of the time, but may not be the best for your images.

Without going into all the technical details – if you are interested in the tech stuff, buy the book I mentioned – I’ll give you an overview of what I have learned so far.

First off, all digital images have inherent noise in them. Since digital images are captured/created by an analogue electronic signal from the camera’s sensor, signal noise is always present. There are a number of factors that can enhance noise. (You know, in retrospect, film seemed so much simpler, but if you have never exposed/developed film, you have no idea of what I’m talking about. I digress.) Using a high ISO setting, above 800, increases image noise. Some older camera, e.g. early Canon 10 D and most point-and-shoot cameras, were known for creating noisy images.

And to make matters even more complex, there are three types of noise – shot noise (that involves the photons hitting the sensor), read noise (that’s when the digital file is created) and dark noise (don’t you love that term – DARK noise – it is created when the sensor heats up when the digital file is created).  The more modern DSLR bodies have overcome the last two types of noise pretty well, but they are still present to some extent.

The thing is, if you sharpen the whole image at one time and you have not reduced any image noise first, all you do is accentuate (sharpen) the noise too, which can really make your image suck.

So the first thing you need to focus on (no pun intended) is reducing image noise. And of course, this is not simple either. There are two types of noise – colour  and luminance noise. Colour noise can appear in the shadows, mid-tones and highlights, but is most visible in the shadows and highlights. Under exposed images have more luminance noise. A properly exposed 1600 ISO image may be less noisy than an underexposed 1oo ISO image.

So what does image noise look like?

Image Noise

As you can see, all the coloured specs do not look all that great. This is the type of image noise that is created by under exposed image, images shot at a high ISO, older DLSR cameras, some point-and-shoot cameras and cell phone cameras. Does this mean all the mentioned cameras produce noisy images? No, but they can.

So the first order of business is to reduce the noise as much as possible before applying any sharpening, otherwise you end up sharpening the image noise and making it look worse.

Image file format impacts image noise too. Jpeg images are more noisy than Raw image files. Most camera makers add in-camera sharpening in their point-and-shoot cameras and for DLSR camera users who always shoot Jpeg images. Now if you don’t enlarge your images beyond say a 4×6 print, or if you only view them online on Facebook, noise may never be an issue,  But if you make a larger print size image noise quickly becomes more obvious.

Photoshop, Camera RAW and Lightroom all have noise reduction tools. Some work better than others, but more on that later. There is also third party noise reduction software available. I have not used and third party software, so I won’t recommend any.

If you read much stuff on the web about image noise and sharpening, you’ll images enlarged a great deal, e.g. 400% + in some cases. You may ask why so much enlargement? This is often referred to as “pixel peeping”. Very few, if any, photographers ever enlarge an image 400% and print it. Theses types of enlargements, in my opinion, are used so that you can “see” what noise and other affects look like when viewing an image or print. Because the ultimate desision is made whe viewing a print. If the print looks good, then who cares how it was made? If the print is less than stellar, then you want to be able to recognise what is wrong and how to avoid it yourself.

That is the main reason I write this Blog. Let me know if you find this helpful. Mind you, either way I’m still writing the Blog…. so there.

Stay tuned for more!

I am currently making my way through a book on sharpening digital images – Image Sharpening with Adobe Photoshop, Camera Raw and Lightroom by Bruce Fraser and Jeff Schewe http://www.chapters.indigo.ca/books/search?keywords=image%20sharpening&pageSize=10 .

You can get it new or used – I chose the used route and was happy with the results. (I also tend to mark up my books, so no one is going to want my copy, but that’s beside the point.)

Based on past experience with Photoshop (a really old version) and Lightroom (most current version) I tend to sharpen my images visually. Or more simply put, I use what I see on the display to evaluate the final image. Apparently, the wrong approach.

I also tended to sharpen the whole images at once and play around with the Radius and Amount buttons again until I like what I see on the display. Again, the wrong approach.

In short, I am doing everything wrong and I’m sure I’m not the only one. I belong to a local camera club – they have a great studio that would cost a fortune to rent and it is free to members – and from talking to other members everyone seems to follow the same approach I do. Of course, the only way to really evaluate sharpening quality is to look at the final print or screen image. (The authors agree with me on that point. Okay maybe it was really their point and I just happen to see the logic of it.) In fact, the authors state that good sharpening for a print can often look really hideous on the display.

Since I am still working my way through the book, I will highlight some of the things I have learned so far that may, or may  not, surprise you and make you think.

First off, the book deals with processing Raw files. They give a very clear description of why working with jpeg files does not work all that well. (I will not add to the debate about RAW vs JPEG.)

Second, there are three things that you need to realize about digital images:

1. the digital camera inherently creates soft image detail and image noise in every capture
2. the wrong kind of sharpening can exaggerate texture we don’t want (noise) and obscure detail we do
3. when printing – translating pixels to ink dots – some softness is introduced

The author’s broad recommendation is multiple pass sharpening by developing your own sharpening workflow. They acknowledge that this is considered heresy – you ONLY sharpen images once! (My goodness, what are they thinking!)

Sharpening is not just about emphasizing detail. It also includes the blurring of details, e.g. to enhance a flower you may blur the background details somewhat and thereby render to flower’s petals as “sharper” when compared to the background. (Now there’s a novel thought.)

An old axiom is if you want sharp images, use a good quality tripod. I learned this from my film days. And it is even more true with my digital camera. In fact, digital cameras were initially rejected by both professionals and the public because the images suffered from camera shake (as a result of called shutter lag). That is not as big an issue any more, but camera shake does creep into a lot of digital captures.

Sharpening cannot fix sloppy focusing. I remember in my film camera days, we often spoke of using “Sharpene” to sharpen out of focus negatives. “Hey just run it through some Sharpene and your neg will be fine.” Yeah, right.

Using your camera’s auto focus (AF) does not guarantee that the image is sharply focused. Higher end cameras, e.g. Canon 5D MKII lets you adjust the micro focus of the lens. You can buy a tool that will let you see exactly how accurate your camera/lens combination is focusing using AF. (Check out http://www.rawworkflow.com/ for their Lens Align product.) (If you don’t want to buy the product, contact me for and $20.00 I’ll test your camera/lens combination and show you how accurate, or not, it is actually focusing using AF. edward@amindseye.com) I digress, sorry.

What exactly is image sharpening? When looking at an image, we judge its sharpness, or lack of, but looking at edges. If the edges of things in the image appear distinct and separate, we deem the overall image to be sharp. It is that transition area that is the key to visual sharpness. If the edges appear too sharp (crunchy), our brain tells us something is wrong. We may not know right away what it is, but the image just looks off.

Image size and viewing distance. The print size also affects our perception of an image’s sharpness. Large prints, i.e. 16×20 or greater, are viewed from a greater distance than say an 8×10 print. We’ll hold an 8×10 print in our hands if allowed. However, in an art gallery a framed 16×20, or larger print, behind glass and a mat and framed is viewed from a distance of about 4 – 5 feet. Our eyes take in the image a whole entity. Depending on the quality of our vision, or when we had our prescription for our glasses updated, we may view an the print as sharp or not. The key point is image sharpness depends on a number of factors other than what you see on your computer display.

Excellent sharpening is invisible. Think about that for a moment. One could say that about any digitally processed image – the processing should be invisible to the viewer. I often hear people viewing images ask, “Was the shot Photoshoped?” The viewer senses there is something wrong but they are not really sure what it is. Photoshop has now become a verb.

This will be a continuing series – one, because I”m not finished reading the whole book yet; two, because there is so much to learn and understand before you even touch your computer’s mouse, this entry could go on forever.

For those who read this, please add a comment if you are interested in hearing more.

My current interest area is the Conversion of 2D images to a 3D models.

I found an interesting website,
http://make3d.stanford.edu/index.html

This website has amazing videos which show what can be done with an image.

I downloaded some pdf documents and trying to understand it.

Though I have not much time but image processing is an ammazing field, and I am enjoying it much more.

A precise cochleostomy is a crucial step in cochlear implantation, particularly if residual hearing is to be preserved. A contactless ablation of the promontory bone by a pulsed CO2 laser system seems to be a promising approach. The bone is removed by a scan head controlled laser beam in sequential scan cycles with a pulse rate of 50-100 s. Digital picture analysis and pattern detection are used to identify the membranous lining of the cochlea. We achieved a bone ablation in a micrometer range per scan cycle with the laser. A perforation of the promontory bone could be detected by automatic pattern detection. The enhancement of automatic pattern detection can lead to a minimally invasive, function-preserving laser cochleostomy. Copyright © 2009 John Wiley & Sons, Ltd.

from Cochlear Implants International

Find out about today’s use of computers in your old grammar school. Remember if and how your teachers included New Media in the classroom. Describe your findings and your opinion about it

Media education plays an important role in my old grammar school. I went to the St.-Ursula-Gymnasium in Attendorn (www.st-ursula-attendorn.de) which is a private school under sponsorship of the archbishopric Paderborn.

The school comes with two computer rooms, with 16 computers each. The students of the Oberstufe have the possibility to access another computer room with 6 computers at any time. In addition, there are 4 computers in the library, which can be used by everyone.  

The school established a general programme including the use of New Media in the different grades of Sekundarstufe I: New Media Programme (General)

Beside this, there are some more projects encouraged in certain subjects,  for example e-mail projects in foreign language classes. By this students get in touch with students from English – French – or Spanish speaking countries (depending on the subject).  The projects are guided by a teacher.

Furthermore, students have the possibility to join an afternoon group called “AG Simulation” in which they make use of the developer software Visual Basic.Net to programme simulations and learning software for the subject history. The part taking of this course is voluntarily. Short Article on “AG Simulation”

Next, I want to give you an example of how computer’s are used in class. In the school’s yearbook of 2007 there is an example given of using Adobe Photoshop for Digital Image Processing in Arts. Student Article on Digital Image Processing in Arts (2 pages)

I can remember that some of my teachers also included New Media in the classroom. Although I don’t remember everything exactly, it is still in my mind that we for example had to create a PowerPoint presentation on Caesar in grade 8 (subject: Latin). Furthermore, we had to do quite a lot of research on the internet for projects and presentations in Geography in the Oberstufe (e.g. to find out in which stage of economic development a country was according to Rostow’s  Model …). Most of the time these searches were unguided. We were free to use PowerPoint for presentations, but I remember that only a few students made use of it. I remember that I never made use of it, because I was afraid to handle it and nobody showed us how to use it properly (the creation of the PowerPoint presentation in grade 8 was only a crash-course …). Sometimes we made use of the programme Euklid in Maths.

What I remember and also liked best was the use of computers in social sciences. The topic of our last term was “socionic”, a mixture of social sciences and computer sciences. Dealing with this topic we made use of Sugarscape, “an artificially intelligent agent-based social simulation”.  These simulations can carry out a great number of different scenarios and are useful to visualize how societies develop under certain conditions.

If you are interested in this visit http://www.sowi.st-ursula-attendorn.de/zw/zw_netz.htm and ask me for the password to download more detailed information. Unfortunately I’m not allowed to publish it without a password because it’s protected by copyright.

I especially liked to work and learn with the sugarscape simulation. It was motivating and facilitated the comprehension of social theories. Additonaly, we got the possibility to create our own sugarscape scenarios after we got into the topic.

So all in all I would say that I got in touch with New Media in school relatively often, although the sequences were never too long (except for the sugarscape simulation in social sciences). I think my former school is still on a good way to school its students to become media competent.

I found the solution of DIP 2/e from internet.

Digital Image Processing 2nd Edition (DIP/2e) by Gonzalez and Woods ; Language: English ; ISBN : 0201180758

Digital Image Processing (2nd Edition)

Anyone who has ever worked on Digital Image Processing is surely acquainted with the image of Lena (aka: “Lenna”):

I doubt there’s one single book on Digital Image Processing that does not contain this image. The Lena photo can be found on countless scientific papers. Numerous algorithms have been tested using her photo. Lena is definitely a celebrity in the Image Processing community.

But, who is the mysterious Lena after all?!? She’s Lena Söderberg (b. 1951), a Swedish model who is also known by her maiden name Lena Sjööblom / Lenna Sjööblom. She appeared in the centerfold of the November 1972 issue of the Playboy magazine. In mid-1973, engineers at the Signal & Image Processing Institute were searching for good test images. Someone found a copy of the November 1972 issue of the Playboy magazine, and the engineers scanned the centerfold photo of Lena. Over the past decades, the Lena image has been the de facto standard for testing Image Processing algorithms. The mysterious Lena is considered the “First Lady of the Internet”, and an “Information Age Madonna”. Apart from Mona Lisa, no image has been studied harder.

In 1988, Lena Söderberg was pleasantly amused to find out what had happened to her picture. In 1997, she was a guest at the 50th annual Conference of the Society for Imaging Science and Technology, where she was busy signing autographs, posing for pictures, and giving a presentation about herself.

__________

Related:

• Culture, Communication, and an Information Age Madonna

• original, uncompressed image of Lena (TIFF – 768 KB)

• the full Lena centerfold photo (Warning: contains nudity)

• Playmate meets geeks who made her a net star

• Imaging Experts Meet Lenna in Person

• The “Lena image” comes to life

• A Complete Story of Lenna

• The Search for Lena

• The Lenna Story

• The Lena story

• A Note on Lena

Hari ini kutemukan sebuah website bagus yang mengilhami topik skripsi mahasiswa TI, sebuah website yang mampu mengubah gambar ke teks dengan pola yang sama dengan gambarnya, tapi berisikan tulisan dengan huruf-huruf seperti yang kita inputkan… dan hebatnya lagi hasil akhirnya adalah gambar lagi…

Jadi menurut saya, ada 2 macam proses yang dilakukan:
1. Gambar ke teks untuk dibentuk pola gambar dengan isi teks yang sesuai inputan
2. Dari teks tersebut ke gambar lagi dengan ukuran tertentu.

Algoritma nya mirip dengan ASCIIArt hanya saja kali ini teksnya berwarna, dan kemudian tentu ada edge detection dan berbagai algoritma citra lain. Lalu pengubahan dari pola teks ke gambar lagi…

Ada yang berminat?? silahkan kunjungi http://textorizer.whatfettle.com/

Contoh gambar hasil:

We offer basic and advanced Digital Image Processing Services and solutions like morphological image processing, photoshop image processing, satellite image processing, color photo processing and vector image processing using latest techniques.

Our strength in providing processing solutions for digital images like image masking, image re-touching, back ground cleaning & cloning, sharpening & restoration of color, all this derive from our vast experience. Our dedicated team of highly skilled and experienced professionals serves the image and photo processing requirements of our clients. An in house quality checking mechanism assures value addition and delivers perfect job. Our team can deliver your job overnight or in time bounded fashion irrespective of job size.

We also offers excellent scanning of images to a format of your choice like an excel spreadsheet or a database of your choice. Our exacting standards ensure superb image capturing, image keying, and image storage and retrieval. We have several tools to handle data conversion of your scanned images and images data entry is catching up as main outsourced service.

Our mainly services are

1. Photo Restoration
2. Photo Enhancement
3. Photo Retouching
4. Vector Conversion
5. Pop Art
6. Web Design

Photo Processing Services

Our Photo processing services includes advanced image cutouts/clippings, balancing brightness / contrast in an image, repair of minor scratches, creases, minor dust, and spot removal to removal of major scratches, cracks, creases, and stains in all areas. We also repair seriously faded/damaged photographs and make it print ready.