Laptops and webcams still lag way behind when it comes to image quality, with many of us spending hours each week, staring at dark, blurry video. Our results improve coloring, brightness, and sharpness of laptop image quality in real-time, and also solve HDR challenges like light coming from windows and casting dark shadows on faces...
In addition, we offer video enhancement features such as bokeh (background blur), background replacement, auto-framing, and gaze correction.
Higher-end smartphones with Night Mode can take decent quality still images in low light, but when it comes to video, even the best phones struggle...
Our mobile results show that with our AI ISP, even low-budget smartphone cameras can beat the market-leading phones when it comes to low light imaging, and HDR scenarios. In addition, we offer more accurate coloring and sharpness.
For applications like CCTV protecting a property, or security cameras in outdoor locations, low-light performance is critical, as crimes and theft usually occur at night. Our technology improves brightness and color to enable high quality, full color, real time images in extremely low light.
But our ISP doesn't just improve imaging at night. Security cameras often face scenarios of high dynamic range, with varying sunlight and shadows. Our software tackles these HDR imaging challenges.
Smart cameras today use object recognition to help tell the difference between, for example, a neighbour's cat and an intruder. In addition to improving video quality, our software improves accuracy of object recognition.
A drone is an expensive investment and if it can operate safely and effectively in all lighting and weather conditions then it's use can be extended by extra hours each morning and evening...
Using AI for image enhancement can solve other challenges for drone vision for either photography or navigation. For example it can be used to eliminate the effects of vibration, wind, rain, or to improve image performance when there is high dynamic range.
The video endoscope uses a tiny image sensor, and some LED lights, to enable surgeons to see, and operate, inside the human body.
Minimizing the illumination is key in the application. The body tissues have liquids and fats and are highly reflective. Our denoiser allows clear images in much lower light levels than conventional technology...
Our software ISP can also be easily tuned for the specific requirements of the application, which in this case would be greater detail and finer differences in the red and yellow parts of the colour spectrum.
The use of our technology will enable endoscopes to have the smallest possible tip size, and the best image quality, leading to faster, more accurate and less invasive surgery.
As regulators seek to reduce the number and severity of road traffic accidents, machine vision cameras are being deployed in the majority of new vehicles worldwide...
Visionary.ai technology can be used for internal cameras for driver alertness monitoring, particularly for night driving when the cabin is very dark. Good quality images of the drivers face and eyes allows algorithms to accurately assess the attention, or drowsiness of the driver.
It can also be used to improve image quality from the front, side or rear facing cameras so that downstream algorithms have a great probability of correctly identifying and responding to hazards.
Whether it's a doorbell, an oven, a phone, washing machine, or vacuum cleaner, almost all our electronic devices now use cameras. Whatever your application, we can help improve image quality...
Our technology can also be used in AR/VR applications, and virtually anywhere else there is a camera.
Cameras will be increasingly used in smart city deployments for applications such as license plate recognition. The ability of visionary.ai to remove noise from images enables high confidence machine reading of vehicle license plates, in all lighting and weather conditions. As the denoiser works in real time it can be deployed in fast moving traffic, or used to drive real time decision making processes.