Welcome to follow my WeChat public account: Xiao Zhang Python!
Hello everyone, I am zeroing~, today I will share a fun website with you, just upload the old local photos, and then the website will automatically color the black and white photos
Old photo color technology is already out very early, when in May last year, a Technology Up Main Station B Otani game creation hut by the Artificial Intelligence 100 years ago (that is, the Qing Dynasty) image of old Beijing were reduced to The tone of the image has changed from black and white to color
Up here as the primary video presentations, video technology is not only used the color, but also used the pictures repair technology;
About picture repair I do not find relatively easy to use and free open-source website to see basically a fee,
Take the above e-commerce website as an example, here are the service prices for repairing and coloring old photos; it is quite good to see the repair effect according to the product display diagram, but the price is a bit expensive (1 HKD equals RMB 0.8), , Picture repair If you can’t find a better tool for the time being, you can only reproduce a paper in a related field and write an article to introduce everyone
The website introduced this time is called Deep AI . It shares some cutting-edge Paper, research results, and information related to Deep Learning. In addition to providing high-quality information media, the website also provides some relevant information such as recruitment positions.
Regarding the introduction information of this website, neither Google nor Baidu has included it, but judging from its Github homepage, it should be formed by a group of Ai fans. [ Use the power of artificial intelligence to promote the development of image and video applications ]
One of the modules in the website is the most interesting and important one in my opinion. It is about the initial landing of the research content; the trained deep learning network is provided with an API call interface on the website, and visitors can easily apply technology to their own data It lowers the threshold for use.
[Deep AI] The API interface provided on the website involves many technologies: image coloring, face recognition, grid migration, super-resolution, image noise reduction, human posture detection, generating images from text, etc...,
There are too many to introduce, and a screenshot ( only part of it ) is posted below :
Here today, a brief two techniques: black and white photos to color and super-resolution ;
Super resolution
Website: https://deepai.org/machine-learning-model/torch-srgan
Super resolution is to increase the resolution of the image through the network, make the blurred image clearer and the original information in the image will not be missing. The effect after the web page is opened is as follows
The website provides two ways to upload pictures: 1. Local upload, click the [Image] button to upload; 2. Url upload, upload through the picture Url connection, click the [ url upload ] button on the page to use the method ;
After the picture is uploaded, the website will automatically process the uploaded picture, and save it after processing. The usage is as follows:
Take a look at the effect after processing. Before super-resolution processing is as follows, the image size is 14.4 KB
After processing, the image size is 745 KB, which is nearly 50 times larger
If you look carefully, the image is indeed clear, but you will find that there are some defects, and some areas of the image will appear distorted and distorted;
Picture coloring
Website: https://deepai.org/machine-learning-model/colorizer
Color pictures to use and super-resolution similar, you can refer super-resolution using the method, only show here before and after handling renderings
Before processing:
After processing:
Before processing:
After processing:
The effect looks good, but if you look closely, the colors on the Internet are all cool colors, blue, purple, and light yellow; when compared with the original black and white pictures, there is no big contrast; relatively bright red, rhubarb, etc. There are few warm colors, and the coloring effect has a lot of room for improvement.
There are a lot of very interesting technologies on the website, I will not introduce them here, and interested friends can study them by themselves;
In addition, in addition to manually uploading pictures, if there are more pictures (or videos) that need to be processed, you can also use language scripts, such as Python, curl, Javascripts, etc.; the specific usage method website also provides documentation, here I post Here is how Python calls the API method:
# Ensure your pyOpenSSL pip package is up to date
# Example posting a image URL:
import requests
r = requests.post(
"https://api.deepai.org/api/colorizer",
data={
'image': 'YOUR_IMAGE_URL',
},
headers={'api-key': 'quickstart-QUdJIGlzIGNvbWluZy4uLi4K'}
)
print(r.json())
# Example posting a local image file:
import requests
r = requests.post(
"https://api.deepai.org/api/colorizer",
files={
'image': open('/path/to/your/file.jpg', 'rb'),
},
headers={'api-key': 'quickstart-QUdJIGlzIGNvbWluZy4uLi4K'}
)
print(r.json())
Okay, the above is the entire content of this article. Finally, thank you for reading!