“Until now, medical device designers were forced to make a tradeoff between size and resolution,” said Tehzeeb Gunja, director of medical marketing at OmniVision. “The dimensions of camera modules and their integrated image sensors are limited by the endoscope’s outer diameter. Additionally, to obtain a sufficiently bright image while limiting the heat dissipation of the LEDs at the distal tip, medical image sensors must have a relatively large pixel size. Almalence’s SuperResolution algorithm enhances the sensor’s resolution and image quality, enabling doctors and nurses to view its captured images on high-definition monitors and tablets.”
“The vast majority of high-end smartphones on the market today use our SuperResolution algorithm because we produce the greatest image quality within the limited space available, by reconstructing details beyond the physical limits of a sensor’s pixel count,” said Eugene Panich, CEO of Almalence Inc. “We have expanded our longstanding partnership with OmniVision in mobile imaging to bring our industry-leading algorithm to medical designers, enabling them to create small-diameter endoscopes with greater detail and better image quality than the image sensor would otherwise be capable of producing.”
Competing solutions include an image signal processor (ISP) that provides some level of upscaling as part of its processing pipeline. The functionality of these solutions ranges from the simple repetition of pixels, resulting in poor sharpness and image definition; to complex, computationally intensive solutions that provide upscaled images with relatively good image quality. However, none of the competing ISP solutions can provide the ability to capture more details than the sensor is capable of producing. Almalence’s SuperResolution algorithm uniquely increases the sensor’s effective resolution by not only making the image look sharper and smoother, but also reconstructing more details, as if it were captured with a higher resolution sensor. Additionally, the algorithm’s computational load is optimized to run on any major SoC or DSP, leaving plenty of headroom for other processing tasks.
OmniVision’s OVM6948 CameraCubeChip module is available now for sampling and mass production. Likewise, Almalence is currently licensing its SuperResolution algorithm, which is pre-tuned to perform optimally with the OVM6948. Additionally, the Almalence license includes expert tuning services for any OmniVision medical image sensor in combination with any lens available on the market—at every stage of development, from initial design to product launch. Almalence also provides a software development kit with its license, featuring numerous parameter APIs for additional tuning by medical device designers.
