Plain human observation of a damaged car can be inaccurate and biased. Developers need an efficient starting point to develop tools for their clients in the field of insurance and mechanics to ease down the process of identifying, locating and assessing car damage.
Nowadays, there are more and more cars in the streets, and the the frequency of car crashes is exponentially growing as well. The demands for insurance companies to cover the damage expenses is shocking. This the need of vehicle damage detectors that can ease the work of adjusters and mechanics.
Vehicle Damage Detector API is an reliable tool that avoids time-consuming observation that´s always inaccurate and ineffective. That is why this new API is furnished for developers and ultimately for their clients in the field of insurance and car fixing, that will help find any damage in an image with no effort at all.
This API is the best tool for images and thoroughly scanning all the vehicle, so as not to miss anyt damage at all. It identifies the intensity of the damage, the location on the car, and even the repair costs, with an efficient apraisal of the damaged parts.
This damage detection API is the best tool to avoid trouble on detecting any kind of damage and it`s perfect for any type of situation. For the ordinary user´s own security, this API will detect any kind of break-in or theft to claim the insurance company. It guarantees that whether it´s a developer, an insurance adjuster or an ordinary user won´t miss any car damage.
How does it work?
This new API works with an AI system that will analyze each pixel of the input image and detect any type of damage, namely cuts, scratches, tears or even spots or stains on the images. By just providing an image, the API will respond with a percentage of the damage and the location and type.
Counting on a subscription on Zyla API Hub marketplace, just start using, connecting and managing APIs. Subscribe to Vehicle Damage Detector API by simply clicking on the button “Start Free Trial”. Then meet the needed endpoint and simply provide the picture of the product. Make the API call by pressing the button “test endpoint” and see the results on display. The AI will process and retrieve an accurate report using this data.
What About Pricing?
This image damage detection API offers other three plans: the Basic plan, which includes 1,000 Requests per Month; the Pro plan, which includes 10,000 Requests per Month; and finally, the ProPlus plan, which includes 100,000 Requests per Month.
It`s advisable to keep the input image under 5MB, between 2000×2000 and 640×480 in resolution for a more accurate response. To avoid longer process times it´s recommended not to use higher resolution images. For best damage detection, it’s recommended to take a close picture of the area, approx. 80cm – 1.5m when taking the picture.
If the input is “draw_result”: true,
“image”: “https://i.insider.com/54da54336bb3f7156eeef89e?width=800&format=jpeg”, the result will look like this:
{
"draw_result": true,
"job_id": "52af644d-03c9-46c8-ab8c-c486fc432ade",
"output": {
"elements": [
{
"bbox": [
372,
145,
489,
245
],
"damage_category": "slight_scratch",
"damage_color": [
50,
50,
100
],
"damage_id": "1",
"damage_location": "rear_bumper",
"score": 0.613146
},
{
"bbox": [
322,
411,
444,
481
],
"damage_category": "slight_scratch",
"damage_color": [
50,
50,
100
],
"damage_id": "1",
"damage_location": "rear_bumper",
"score": 0.602163
},
{
"bbox": [
403,
300,
487,
391
],
"damage_category": "slight_scratch",
"damage_color": [
50,
50,
100
],
"damage_id": "1",
"damage_location": "right_rear_tire",
"score": 0.422464
},
{
"bbox": [
404,
311,
491,
383
],
"damage_category": "severe_scratch",
"damage_color": [
0,
0,
120
],
"damage_id": "2",
"damage_location": "right_rear_tire",
"score": 0.497413
}
]
},
"output_url": "https://vpai-demo-portal-shanghai.oss-accelerate.aliyuncs.com/vehicle-damage%2Fdraw-134451.7685_20220804134451.jpg?OSSAccessKeyId=LTAI5tGgUWzx5gZopTcAvNtp&Expires=1659620752&Signature=V3JNzNARJx8PvRclb1DB%2Bb4Dgw8%3D",
"url_expiry": "2022/08/04 13:45:52"
}