I want to create an automated (python) license plate generator script. It needs to work as follows:
1) get license plate template (image of license plate with no licence plate characters on it) [this should be an input option]
2) get license plate fonts (see [login to view URL])
3) get font size (width & height) and exact location (x & y coordinates from top left of license plate) of license plate characters based on selected country/state and provide this as x, y, w, h for each license plate character (in json file)
4) generate random license plate number onto he template using above information
5) apply a perspective transform of license plate by random roll, pitch yaw as well as scaling. This is to emulate different view angles and distances
6) obtain a number of background images and size them to the same size (you can use images here for ex [login to view URL])
7) randomly place the rotated & scaled license plate image into the background image
8) apply the following types of noise on the resulting image:
a) Gaussian blur, Poisson noise
b) motion blur
c) resolution blur (pixelation)
d) lighting adjustment (saturation, gamma, contrast)
e) [bonus] mist, rain, rain droplets on windshield
9) output resulting image; also save exact location of 4 corners of rotated/scaled license plate, all the noise parameters used, and the license plate number itself
Here is an example code to start: [login to view URL]
also see this blog post: [login to view URL]
17 pekerja bebas membida secara purata $674 untuk pekerjaan ini
Hello. I'll be happy to do that using OpenCV :) Should we just place it on background without trying to tie it to something? Let's talk and do the business!
Hi, How are you? If you trust me ,I 'll do my best for you and my reputation. Please let's discuss more detail in this chat. Thanks for your attention.
Please join me for project scope discussion . Waiting for your message . I shall provide you free 1 year support after the delivery also. I can create games , website, Android, IOS and desktop application+ graphics .