Image processing methods often introduce distortions, which affect the way an image is
subjectively perceived by a human observer. To avoid inconvenient subjective tests in cases …

为提升激光雷达点云目标识别的效率和减少计算开销, 本文通过采用改进的邻域确定方法和LRF
估计方法, 提出了一种新的特征描述子: 半球单值形状上下文特征描述子(Hemispheric Unique …

Feature detection plays a crucial role in image registration. There exists quite a few feature
detection algorithms in literature like BRISK, FAST, SURF etc [14]. Each of these algorithms …

To address the scale invariance and the mismatching problems of ORB (oriented FAST and
rotated BRIEF), an improved algorithm based on multipurification was put forward and …

Keypoint detection is a basic step in many computer vision algorithms aimed at recognition
of objects, automatic navigation and analysis of biomedical images. Successful …

In environmental perception of outdoor scenes, the number and repetition rate of feature
points extracted based on fixed threshold change dramatically with the change of …

Real-time object tracking systems must be capable of tracking an object subject to abrupt
changes in motion and appearance. Model uncertainty is inherent in the tracking process …

In today's world, there is a continuous need for automated appliances. With the increase in
the living standards, there is an immediate need for developing circuits that would change …

Feature extraction and object matching is one of the important and significant task in many
computer vision and robotic applications. Finding the corresponding points between the two …