In binocular stereo vision, stereo matching is a critical technique that has always been the focus and challenge of research in the field of stereovision. The results of stereo matching directly impact the effectiveness of 3D reconstruction. Due to the limitations of traditional region-based local stereo matching algorithms, which rely solely on similarity measure functions using information such as grayscale, color, and gradient from a point's neighborhood to compute matching costs, these algorithms have lower computational complexity. This can lead to problems such as incorrect matching in areas with repetitive textures, weak textures, and depth discontinuities. Therefore, building upon traditional region-based local stereo matching algorithms, this study investigates a global stereo matching approach based on dynamic programming. This method employs global constraint information from the image and constructs a global energy function for matching costs. Leveraging the principles of dynamic programming and stereo rectification, the approach decomposes the process of solving the entire image disparity into several subprocesses, solving them sequentially. The specific process involves adhering to the order of epipolar lines, seeking the minimum cost path on the disparity image, and obtaining the final disparity map. As the global stereo matching algorithm based on dynamic programming considers the disparity constraints between pixel points along epipolar lines, it can effectively address mismatching in depth-discontinuous areas and regions with uniform textures, yielding favorable outcomes.
One of the key components of a belt conveyor, the conveyor belt, often operates in challenging environments and is prone to longitudinal tearing faults. In order to promptly detect longitudinal tearing faults in conveyor belts and minimize the losses caused by such failures, this paper proposes a visual defect detection system assisted by line laser technology. This system employs an industrial camera to capture images of the conveyor belt's underside illuminated by a line laser. Initially, the images are preprocessed, and a hierarchical pyramid-shaped model with a level of 1 is established for matching purposes. A portion of images depicting conveyor belts without longitudinal tearing is used for training to generate a shape template. Once the template is created, it is employed to detect anomalies in the images under examination. When a comparison between the test image and the template surpasses a predefined threshold, anomalies are identified and marked, allowing for an effective determination of whether the conveyor belt is torn.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.