WiMi Hologram Cloud Inc., a leading global Hologram Augmented Reality (“AR”) Technology provider, announced that a new method based on characteristic spatial object (CSO) extraction and matching is developed. First, the CSOs and their localization points on the image are automatically extracted using the Mask R-CNN model. Then, each object and its nearest neighboring objects are coded according to categories, relative distances and relative directions by a coding method. Next, a code matching algorithm is applied to search for the most similar object pairs. Finally, the object pairs are filtered by position matching to construct the final control points for automatic image matching. The result of the experiment showed that this method outperforms the traditional method in terms of matching success rate.
With the continuous development of remote sensing technology, automatic matching of high-resolution remote sensing images (HRRSIs) has been a challenge. Different angles and lighting conditions can lead to local distortion of images, which brings troubles to data processing and analysis. WiMi’s automatic matching of high-resolution images based on CSO is a breakthrough technology that will revolutionize the remote sensing industry, realize accurate data analysis, and provide reliable information support for environmental monitoring, urban planning, and agricultural management, among other fields.
The core of this technological innovation is an automated matching method based on CSOs. Traditional image alignment methods usually rely on grayscale alignment, transform domain alignment, or feature point-based matching, but these methods are very sensitive to grayscale, rotation, and distortion, and are computationally huge, making them unsuitable for automated matching. This technique adopts a new idea to achieve more accurate matching results by automatically extracting CSOs and locating their positions using the Mask R-CNN model.
First, the image is scanned using the Mask R-CNN model to automatically extract the CSOs and their localization points on the image. The accuracy and efficiency of this step are based on the team’s years of accumulated research and innovation in computer vision. Subsequently, each extracted CSO and its nearest neighboring objects are encoded based on object class, relative distance and relative orientation. The encoded features provide the basis for subsequent matching.
Also Read: WiMi Developed Metasurface Eyepiece for Augmented Reality with Ultra-wide FOV
To find the most similar objects, WiMi uses an advanced code matching algorithm. The algorithm determines the degree of the match by calculating the similarity between the encoded features. Object pairs with higher similarity are considered candidates for matching. Further, the initial object pairs are filtered using the position matching algorithm to exclude some false matches and obtain more reliable alignment results. Through this step, WiMi’s technique is able to accurately find the spatial location links in the image and further improve the accuracy of the matching.
WiMi’s automatic matching for high-resolution images based on CSOs has achieved impressive results in experiments. Tested and compared across multiple datasets, the results show that the technique significantly outperforms traditional local feature point-based optimization methods in terms of matching success rate. This breakthrough result will enable the remote sensing industry to process and analyze data more accurately and efficiently, providing a more reliable basis for decision-making.
In addition to remote sensing image processing, the technique has a wide range of application prospects. In the field of urban planning, the automatic matching technology based on CSOs can help planners better understand the development trends of cities, so as to formulate more scientific urban development strategies. In environmental monitoring, the technique can provide accurate image matching results to help scientists monitor and assess environmental changes, providing important data support for environmental protection and resource management. And the technique can also play an important role in agricultural management and disaster monitoring, providing accurate data analysis and decision-making support.
WiMi is bringing the technology to market and exploring its applications with industry partners. By integrating the technology with existing remote sensing data processing platforms and software, users will be able to easily realize high-precision image matching, thus improving the accuracy and efficiency of data analysis. The launch of this technology marks another important breakthrough for WiMi in the field of image processing. The application of this technology will bring a great impact on the remote sensing industry. In the past, image alignment required a lot of time and labor, and the results were not always accurate. However, WiMi’s technology will make the matching process more automated, efficient and accurate, greatly improving the efficiency and quality of remote sensing data processing.
SOURCE: PRNewswire
