How to Create Cartoon Effect | Vector Art a lot of photos with the the batch processing.

AFAIK, there is no "magic" way to achieve that effect. The process is very labor intensive.
You can try automatically applying filters from Filter Gallery, but the results will be very hit and miss and nowhere near the quality of the image from that video.

HTH,
Dan

I don't know of any that produce vector art, but I have seen some that do bitmaps.  Googling will find some for you, for example this one that I played with:

Convert to Cartoon - Convert photo to Cartoon - Cartoonize Yourself

---

»bp

If you are using photoshop, the first step is to figure out an easy method to achieve what you want.  If you want to take a lot of time, as mentioned, it can be a tedious process. If you are looking for something quick and not too worried about the "best" end result, you can find any number of quick and easy methods, here is one; http://www.photoshopcstutorial.com/photo-effects-tutorials/cartoon-drawing-tutorial.php.

Once you have desired method you are happy with, you can automate in two ways.  One is to create a folder where you save an image to a folder and the image is processed automatically. This is called a droplet.  Or you can create a batch in photoshop to run manually https://helpx.adobe.com/photoshop/using/processing-batch-files.html.  

In either case, you need to create an action to perform. There are some already made up, but you will want a custom action based on step 1.  https://helpx.adobe.com/photoshop/using/playing-actions.html and specifically recording an action  https://helpx.adobe.com/photoshop/how-to/photoshop-record-action.html

Now that you have an action, when you bring up your batch wizard in photoshop, you simply select the new custom action you created.

Are u trying to tell me that if I follow this tutorial :

https://www.youtube.com/watch?v=VuQhhg_y22E 

and I convert a single photo into an artistic picture like that,with the actions I can give to the program a lot of photos and it will do the same operations that I've done with the first one ? And what happens if,for example,there are a large number of photos without people inside ? To make a good work the script should at least detect if there are human figures or not...

Are u trying to tell me that if I follow this tutorial :

No, that would not be easily automated.  I am sure it could be done with some complex programming, for the purpose of this question, that is not feasible.

What I am suggesting is to come up with a generic option you can apply to all things such as the tutorial I pointed to that is basically:

• Mode - RGB Color
• Poster Edges filter. To do this, go to Filter > Artistic > Poster Edges.
• Adjust the Brightness & Contrast with the values: Brightness Value: 10 Contrast: 30.
• Go to Filter > Artistic > Cut Out.

Use that scenario to create your action.  Before you do them all, experiment with some random images with both people and without.  Play with different values to see what works the best. Search out other simple options.  Like I said, this way will not give you the most optimum results for ALL images, but it will be a good inexpensive option for bulk.

If this is something that requires the highest quality and individual attention, then some more intricate scripting perhaps using machine learning will be needed. There may be some software packages that already have this ability.  You have to weigh the cost/benefit for each option.  

Another option is to use a 3rd party service or api. Perhaps http://funny.pho.to/cartoon/ -> http://developers.pho.to/

Reading
https://towardsdatascience.com/background-removal-with-deep-learning-c4f2104b3157
https://www.zdnet.com/article/how-microsofts-caricature-ai-turns-photos-into-portrait-cartoons/
http://www.shellandslate.com/fastmedian.html
https://www.kapwing.com/cartoonify

3rd party apps
https://www.befunky.com/features/photo-to-cartoon/
http://www.cartoonize.net/
http://www.itoon.net/itoon.htm
https://www194.lunapic.com/editor/?action=cartoon
https://www.befunky.com/features/photo-to-cartoon/

Command Line Photo Manipulation
https://www.imagemagick.org/ 
https://imagemagick.org/discourse-server/viewtopic.php?t=31416

Hello,

I'm completely fell in love with the graphic style of this music video :

https://www.youtube.com/watch?v=ss3NhlKY86M

and I would like to know if I can achieve it by using a plugin,a filter or a combination of effects and batching procedures. thanks.

anyone again for this question ?

I have provided you with a lot of details, what have you tried so far? what worked? what triped you up?  please provide details and your before and after image.

@Scott: I don't think we're helping the asker. We're wasting his time.
The two youtube videos that he likes involve a lot of skill, talent and time. The asker wants to know if he can substitute the skill and talent with a button/photoshop action.
"You can't do that" seems to be the right answer.

the point is that I don't understand what "requires skill and talent" and what "can be done with a button / photoshop action". What I can do is to explain which kind of graphic style I like more.

Mario,

Your original question was asking about a tool to create a cartoon effect with the ability to batch process 100,000 images.  This part I think we have answered.  In short, there are 2 routes. 1) Use advanced tools and scripting or 2) Use an API or other 3rd party service.  Both of those options have been detailed.

In your recent post, you are giving us an example that you like. That gets into using a specific tool and would be best asked once you have decided which route to take.

The easiest route will be using a third party service because you will have the least amount of controls and it will give you a generic effect with perhaps some limited (but potentially good) options.  This will be the least expensive in terms of your own time and probably the most expensive in terms of spending money.  Here I am talking about simply uploading your entire batch and put it in the hands of somebody else.

The services that provide an API will give you some more choices but you can still spend more time on scripting your batch to work with the api and use their built in filters with limited adjustments and find the best route for all.  This route will cost a little more than the first option in terms of your own time with a potential savings in money you pay out.

The advanced options we outlined such as creating a batch with photoshop will probably yield the best results if you are very picky. It will be the most cost of with regards to your own time and the only cash outlay will be for software.   With this route, I would take a random sample and do some one off adjustments in the tool of your choice (such as photoshop) and play with the options to get the desired results.  See if you can duplicate what you did on your first random photo to the 2nd, 3rd and 4th random photos to achieve a desired result. With the amount of photos you have, this will be more complex to achieve if you are going after the best results possible. The reason is there are going to be different colors uses, different types of backgrounds, different quality of photos, different types of objects in the photos.  You would want to incorporate some AI to first sort out images and apply a different algorithm for type 1 vs type 2 vs type 3 etc.  Then run your batch probably first in smaller sets and test randomly for desired output and keep tweaking your algo to get the best output.  As you can see, this will be very time intensive.  

If you are not very concerned with the quality of ouput, you can choose one method (and I given you a link for an easy method) and batch that for everything.  Keep in mind, that doing this will not give you a consistent look and color output.

We have detailed all of this out.  Now it is up to you to review all of this, do some testing and come up with an option.  For the advanced option, you do need to have skills in coding batch processing, ai and photo manipulation.    

If you had only one image, you could go into photoshop and as the instructions point out in the link I provided, just play with hue and saturation levels until you hit something you like. You just can't feasibly do that with 100,000 images. If you did, and could turn around an image in 3 minutes that would come out to 5,000 hours.  It would take 10 people working full time about 15 weeks to complete. Or on your own over 2.5 years.   This is what adds to the complexity and not just a question of, "How do I take one specific image and get a very specific output".  And this is why you will need the layer of AI to help sort out the different images in your scripting. Perhaps you can get a good chunk done in a short amount of time and after some QC, you find some images that need help and keep refining for those.  Perhaps then it can be done inside a month or more to get everything to your desired output.

I hope this is making better sense.

the point is that I don't understand what "requires skill and talent" and what "can be done with a button / photoshop action". What I can do is to explain which kind of graphic style I like more.

Hi Mario!

As an illustrator, I can tell you that most of what you desire is very much a skill/talent versus a blanket effect. In the first video, the artist is performing most of those actions by hand.

Scott provided excellent resources in achieving automated results. I often use Actions and Scripts myself to batch edit. I typically have to experiment with an image first to find the right combination of filters and effects to give me what I want. While doing this, I record the Actions I'm making, save, and then apply the action to a folder of similar images via File/Scripts.

The problem lies in this... when the images you want to edit vary in contrast or brightness.... the script will not be able to know that one darker image may need extra editing (it only repeats what you recorded). In the end, you won't be able to get the same look without extensive modifications to make the images similar enough for the script to perform a consistent styling.

In regards to the styling that you desire in the second video link, your best bet is to use one of the batch methods prescribed by Scott after finding the combination of actions that work best for you. I suggest playing with:

• Image/Adjustment/Levels
• Filter/Filter Gallery/Artistic/Poster Edges
• Image/Adjustment/Hue/Saturation
• Find a free grunge texture ("free grunge texture" will give you a lot of options in Google search)
• Place texture on top of your artwork using the layer blending mode "Multiply"
• https://www.youtube.com/watch?v=v1cwEzXy6Hk

Edit: This would mean raster results, but if you still need a vector for resizing reasons, Illustrator's Live Trace feature may suit you: https://helpx.adobe.com/illustrator/using/image-trace.html