MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202621064727 A) filed by Dr. Manisha Vivek Mane; Dr. Anand Bhaskar; Dr. Asha Durafe; Dr. Jalpa Mehta; Ms. Mamta Tikaria; Ms. Pramila Prashant Shinde; and Mr. Vivek Mane on May 22, 2026, for System And Method For Saliency Detection In Frequency Domain For The Omnidirectional Images.
Inventors include Dr. Manisha Vivek Mane; Dr. Anand Bhaskar; Dr. Asha Durafe; Dr. Jalpa Mehta; Ms. Mamta Tikaria; Ms. Pramila Prashant Shinde; and Mr. Vivek Mane.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The present invention generally relates to a system and method for saliency detection in frequency domain for the omnidirectional images. The invention captures a circular omnidirectional image through a catadioptric vision system comprising a convex mirror and optical lens arrangement, and converts the captured image into a two-dimensional panoramic representation using an equirectangular projection technique. The panoramic image is enhanced through non-linear stretching to compensate for fisheye distortion and non- uniform spatial resolution distribution. The enhanced image is decomposed into multiple frequency-domain sub-bands using a multi- level Discrete Wavelet Transform to extract edge, texture, and boundary information from horizontal, vertical, and diagonal frequency components. Saliency features are computed from the extracted frequency-domain information to generate a saliency map identifying visually significant regions within a 360-degree field of view. The invention improves saliency detection accuracy in omnidirectional imaging environments and is applicable in robotic vision, surveillance, autonomous navigation, and virtual reality systems.
Disclaimer: Curated by HT Syndication.
