I am using http://achecker.ca to validate my HTML/CSS and verify that I am WCAG 2.0 compliant. I'm quite happy with it for the most part but it is giving me an inappropriate error that is frustrating my efforts to get a clean validation. I was wondering if anyone had a suggestion for getting around that?
It's complaining about the contrast of some text I am displaying, saying that the background color and the text color are too similar. That's actually true - background-color is #f2eecf and text color is #f2f2f2 - but it isn't really relevant because I have a photograph appearing above the background color so there is actually good contrast between the photograph and the text color. Also, I've added a 1 pixel black text-shadow around the text to make it even clearer. Unfortunately, the tool doesn't seem to take the shadow or photograph into account. (It makes sense that it wouldn't know how to handle the photograph since it is not a uniform colour but I'm disappointed that the text shadow doesn't help.)
Is there something I can do to my HTML or CSS to help satisfy the tool? The website doesn't seem to have any way to contact the people operating the website or I would have asked there first.
Should I just ignore the "errors"? Or should I use a different tool? I'm just getting started with WCAG compliance checking - I realize now that I'm rather late getting into this - so maybe there are better tools to use. If so, I would appreciate any recommendation, especially for free tools.
I certainly want my pages to be WCAG compliant but I can't spare money for tools right now.
I have a photograph appearing above the background color
So your text has a transparent background color or it has a background-image? A code snippet would help.
I'm guessing you're getting this error: https://achecker.ca/checker/suggestion.php?id=301
If you look up "achecker.ca" on https://www.whois.net/, you get:
Administrative contact:
Name: Iris Neher
Email: ineher#ocad.ca
If you're using a decent accessibility checker, then you should not try to fake out the checker. I have not used achecker so I can't comment on how accurate it is. But if you post your code example, I might be able to tell if it's really a problem.
You can also try
http://www.brandwood.com/a11y/ - which specifically checks text on images.
https://chrome.google.com/webstore/detail/color-contrast-analyzer/dagdlcijhfbmgkjokkjicnnfimlebcll?hl=en - chrome plugin
Short Answer:
Best practices is that if you are going to composite text over an image, the image is in a container with that container having a background similar to the image.
Longer Answer:
There are a variety of reasons that an image will not load. A site doing things such as placing live text over an image needs to ensure that text remains readable regardless of the load state of the image.
Also keep in mind that a user must be able to zoom up the text in size without breaking the site, meaning keeping the text over the image or BG so it is still readable, etc.
Outlines & Shadows
As for outlines and shadows: while they may assist contrast, they should be used to enhance and not for the sole basis of contrast. In part because the rendering of these properties is far from consistent among different browsers.
Resources
WAVE Automated Checker
APCA Contrast — the new contrast method for WCAG 3
I'm completely new to Tesseract OCR. This problem might be simple but I can't seem to find the answer using Google.
Basically, I have an image that contains two parts: the first part, which is at the top of the image, has a black background with texts in white color; the second part, which is at the bottom of the image, has white background with texts in black color.
I ran tesseract on the image, which correctly recognized all characters in the bottom part, but none in the top part. I am sure that the characters on the top part is very clear and should be easy to recognize by Tesseract. The only difference is that it has black background.
Is there a way to use Tesseract to recognize texts in both black and white background at the same time?
A paper by T. Kasar, J. Kumar, and A. G. Ramakrishnan describes one solution to the problem: "Font and Background Color Independent Text Binarization". The paper can be found here. There is an implementation of the algorithm by Jason Funk. His implementation can be found here.
I have had some success with the algorithm. I think this type of solution is what you are looking for.
You might also find it helpful to review this recently asked question on background removal (OpenCV for OCR: How to compute thresholding levels for gray image OCR) and its answer. You may be able separate regions of interest by background color and then hand each region to tesseract for processing. Alternatively, post binarization you could invert the 8x8 pixel regions (described in answer above) in the black background portion of the image (or vice versus) to create a uniform background.
Finally, you may find some useful information by searching for solutions to the number plate recognition problem (or license plates). Many number plates (license plates) have background images or lighting artifacts that can interfere with recognition. The more general problem is background removal.
I'm talking about the icon set that Github uses for their own site design.
Example Page
On this page, near the top right corner, the "Watch" and "Fork" buttons, as well as the buttons next to them have beautiful icons!
Any idea where the full set can be found?
To clarify, I'm just looking for minimalist grey icons of that size and style to be used in web apps.
I am unsure if the icons Github utilizes are under creative commons. However, the icon packages below are used in other popular software packages that might be a fit for your project (e.g. you can always convert to grayscale and resize them using Gimp):
famfamfam icons
fugue icons
mini icons 2 (already grey for you)
They use wireframe mono icons. You can find them here.
http://www.iconfinder.com/search/4/?q=iconset:cc_mono_icon_set
FWIW, I think you can find them here... https://github.com/styleguide/css/7.0
The directories seem from famfamfam... But the other answer already told you that.
It seems like there is an increasingly popular trend to style hyperlinks in a color that's barely distinguishable from body text. I noticed this just the other day on an SFGate blog page. Also notice that link just a few words ago. Are my eyes getting worse, or is that hard to spot?
I certainly understand styling hyperlinks to look better than browser default styles. But, if they're not easy to see, what's the point of having them in the first place?
My best guess is that designers (or whomever makes the styling decisions) are wary of interrupting eye tracking with colors or other styling that is significantly different than the body text. That would make some sense, but I feel that there's a lot more room for compromise - i.e., styling links to be different enough from the body text that they're easy to spot, while not making them so flashy that they attract the eye to the detriment of easy reading.
Would anyone make the argument that subtle hyperlinks are more effective than more obvious ones? Or, can you point me to any theories or testing conclusions that may justify their use?
#Mike Daniels: Thanks for clarifying that the color I'm having trouble differentiating on this site is the visited link color - I hadn't noticed that, but you're right. The unvisited link color is rather prominent, and it makes more sense to me that visited links would blend in more with the surrounding text.
On another note, I'm not sure that I buy the argument that a hover color/effect is enough to properly distinguish a hyperlink. I don't feel that it serves the cause of usability that a user should have to hover over a link to confirm that it is in fact clickable. I think that should be made obvious at a glance.
I actually have a very good IPS-panel monitor and near-perfect vision with my glasses on. I can see the visited links on this page and those on the SFGate blog if I scan for them, but my argument is that it would only take an underline, a different color, or another visual distinction to make the links stand out much more.
What I'm really wondering is why - assuming that the designer(s) on a high-traffic site like SFGate know what they're doing and have made conscious decisions about link colors - they would choose to style the links to resemble the body text so closely? Is there reasoning behind that?
I can see the reasoning behind not wanting to have hyperlink styles which strongly clash with the regular text, as they can be distracting while reading. But I think that's rarely a problem. If you look at a site like Wikipedia, the links are very visually distinct from regular text, but it doesn't at all make the text harder to read
So the only real reason to only subtly differentiate inline links from standard text is for aesthetic purposes. I think that's what the SFGate designer has done. The link colors they use are easily differentiable when they are not inline (e.g. the Recent Entries or Categories panes), but they are perhaps harder to pick out in the text.
Since they've already carefully chosen a good color scheme that works aesthetically, and you generally don't want to employ too many different colors in a design or use different-colored links in different places, there's not a lot of options. They could have:
bolded the links
underlined them
used dotted underlines (a nice compromise)
or use rollovers
All of which would help to differentiate links from text without needing to give the links a brighter color.
But you also have to take into account the type of page it is and the usage pattern of visitors. If you have a lot of inline links, or if users are mainly there to read an article, not follow embedded links, then you don't want the links to draw so much attention. The SFGate blog is more akin to a digital newspaper. It's not a standard news blog where links are the primary content, or Wikipedia, where the embedded links are also a major focus of the site. Basically, the few links in the blog posts are only there to provide a little supplementary info for the curious but are expected to be ignored by most readers. And they do stand out enough so that while reading the article, you will see the links as you come across them.
The main navbar is readable , and the sub-navbar is normal text with a while background giving the main navbar visual hierarchy. The page is actually alright.
It depends on what you're trying to accomplish. As a site owner, I might realize (via A/B testing) that a subtler link color increases CTR, reduces bounce rate, etc. .
I do, however, think that differentiating your links only by color might cause some problems for colorblind folks.
ok, in SO the visited hyperlink colour is a bit darker and is harder to distinguish from the unvisited hyperlink colour. But I think inline hyperlink design is a tough question to answer. I mean you want someone to know the hyperlink is there but at the same time, is it more or less important than your content? if its the ultimate destination then maybe making it stand out more (i.e. by underlining it) if its evidence for a point and therefore less important that someone clicks it theres no need. I'd argue a simple colour change would do, maybe make the visited/unvisited links a closer match so that they maintain their visibility in the content.
I didn't realize until Mike Daniels pointed it out that your issue is with visited links, not normally-colored links. Visited links have traditionally blended in more, it's not really a new concept. I think the theory is if you've already visited the link you must know it's there, so there's no reason to draw your attention to it anymore; there are other links you haven't visited yet that you're likely more interested in
I can clearly distinguish the link color from the text color. Maybe you should adjust your display settings? I suppose it could be a bit bolder to distinguish it from the surrounding text, but it's ballpark.
It would be nice if the SFGate blog page page used the hover property to provide another visual cue that the thing that I think is a link is actually a link.
Could be one of 2 scenarios, both of which I have experienced.
the design agency isn't web savvy and thinks that different coloured texts and (worse) underlines are unsightly and has asked the designer to make the links visibly uniform (you'd be surprised how often I get asked to do this).
the text has been SEO'd incorrectly by someone who's about 6 years behind the times and believes that you have to link every instance of a term on a page but with a consideration for the visitor, so they've been styled down.
I'm curious if, similar to something Chesterfield said above, someone is under the delusion that they should link lots of stuff on the page in order to improve their search engine rankings. And, because of that, they thought the page might look less ugly if they toned down the coloring of links so that they're more to game Google than for end users.
I'd be curious to see some stats for the number of links per article for under-emphasized links versus articles where the links are clear. Obviously it wouldn't be 100% scientific, but it might lead to more clarity on the issue.
maybe this could be a nice soloution
a {
font: inherit !imporant;
text-decoration: inherit !imporant;
color: inherit !imporant;
}
I was looking a component to reduce red eye effect on taken photos. Ive found an image Processing library, but it does not work well at all. I was thinking on brush an image, manually, and only paint if the base pixel color its red or near in RGB.
Have no time now to explain all the process, but i need help to get this idea, or to get a working comp
Thanks in advance!
If you are wishing to manipulate the pixels of an image/component then pixelbender would be a pretty good way to go.
Have a look at http://www.adobe.com/cfusion/exchange/index.cfm?event=productHome&exc=26 for downloadable examples.
These can be run using the toolkit found at http://labs.adobe.com/technologies/pixelbender/
Video tutorials on pixel bender can be found at http://www.gotoandlearn.com/
With this you could change just the required pixel colours to anything and it works with everything in flash (images, canvas, button's etc)