Computer Vision Techniques

Researchers explore computer vision techniques to enable machines to interpret and understand visual information from the world. This work involves developing algorithms and models that can process images and videos for tasks such as object recognition and detection, scene understanding, image segmentation, and tracking. Specific areas of investigation include 3D object detection from various sensor inputs, including monocular cameras and LiDAR, and applying these methods to complex scenarios like semantic segmentation in remote sensing data.

This research holds significant relevance for Arkansas's economy and natural resources. Applications in agriculture, a key state industry, involve using computer vision for crop monitoring, yield prediction, and precision farming. Advancements in this area can also support infrastructure development and maintenance through automated inspection and monitoring systems. Furthermore, computer vision contributes to public safety and emergency response by enabling advanced surveillance and analysis of visual data.

This research area intersects with machine learning, natural language processing, and robotics. Investigations into advanced neural network applications and sensor-based localization provide foundational and complementary techniques. Engagement spans multiple institutions across Arkansas, fostering collaborative opportunities.