• Journal of the Korea Institute of Military Science and Technology
• /
• v.25 no.4
• /
• pp.381-389
• /
• 2022

Detecting partially occluded objects is difficult due to the appearances and shapes of occluders are highly variable. These variabilities lead to challenges of localizing accurate bounding box or classifying objects with visible object parts. To address these problems, we propose a two-stage part-based attention approach for robust object detection under partial occlusion. First, our part attention network(PAN) captures the important object parts and then it is used to generate weighted object features. Based on the weighted features, the re-weighted object features are produced by our reinforced PAN(RPAN). Experiments are performed on our collected military vehicle dataset and synthetic occlusion dataset. Our method outperforms the baselines and demonstrates the robustness of detecting objects under partial occlusion.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.05a
• /
• pp.101-103
• /
• 2016

Recently, the target object can be represented as sparse coefficient vector in visual tracking. Due to this reason, exploitation of the compressibility in the transform domain by using L1 minimization is needed. Further, L1 minimization is proposed to handle the occlusion problem in visual tracking, since tracking failures mostly are caused by occlusion. Furthermore, there is a weighted parameter in L1 minimization that influences the result of this minimization. In this paper, this parameter is analyzed for occlusion problem in visual tracking. Several coefficients that derived from median value of the target object, mean value of the arget object, the standard deviation of the target object are, 0, 0.1, and 0.01 are used as weighted parameter of L1 minimization. Based on the experimental results, the value which is equal to 0.1 is suggested as weighted parameter of L1 minimization, due to achieved the best result of success rate and precision performance parameter. Both of these performance parameters are based on one pass evaluation (OPE).

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.4 no.3
• /
• pp.158-166
• /
• 2011

Generally, moving object tracking used Lucas-Kanade feature tracking method which is strong in movement, rotation and size. But this method is very weak of occlusion by background or another object and so on. In this case, this method tracks backgrounds or another objects instead a moving object, or a tracking is finished. In order to solve this problem, we proposes Lucas-Kanade feature tracking method which introduce a destimation function and prediction function.

• Journal of Information Processing Systems
• /
• v.14 no.6
• /
• pp.1307-1317
• /
• 2018

To deal with the problems of occlusion, pose variations and illumination changes in the object tracking system, a regression model weighted multi-templates mean-shift (MS) algorithm is proposed in this paper. Target templates and occlusion templates are extracted to compose a multi-templates set. Then, the MS algorithm is applied to the multi-templates set for obtaining the candidate areas. Moreover, a regression model is trained to estimate the Bhattacharyya coefficients between the templates and candidate areas. Finally, the geometric center of the tracked areas is considered as the object's position. The proposed algorithm is evaluated on several classical videos. The experimental results show that the regression model weighted multi-templates MS algorithm can track an object accurately in terms of occlusion, illumination changes and pose variations.

• The Journal of the Korea Contents Association
• /
• v.18 no.3
• /
• pp.11-18
• /
• 2018

To interact with a virtual object in immersive virtual environment, the target object should be selected quickly and accurately. Conventional 3D ray casting method using a direction of user's hand or head allows the user to select an object quickly. However, accuracy problem occurs when selecting an object using conventional methods among occlusion of objects. In this paper, we propose a region of interest based selection method that enables to select an object among occlusion of objects using a combination of gaze tracking and hand gesture recognition. When a user looks at a group of occlusion of objects, the proposed method recognizes user's gaze input, and then region of interest is set by gaze input. If the user wants to select an object among them, the user gives an activation hand gesture. Then, the proposed system relocates and visualizes all objects on a virtual active window. The user can select an object by a selecting hand gesture. Our experiment verified that the user can select an object correctly and accurately.

• Journal of Korea Multimedia Society
• /
• v.12 no.12
• /
• pp.1745-1754
• /
• 2009

PTZ (Pan-Tilt-Zoom) camera based surveillance systems are enlarging their field of application due to their wide observable area. We aimed to detect both static and moving objects in automated working space by using a PTZ camera. For object detection we used background difference method because of the high quality segmentation. However, the method has a problem called 'hole' that is caused by non-continuous surveillance of the PTZ camera and its own characteristics. Moreover, the occlusion which occurs when the moving object overlaps with the static object should be solved for robust object detection. In this paper, we suggest a region-based technique for updating background images thereby overcoming the hole and occlusion problem. Through experiments with real scenes, it was verified that meaningful static and/or moving objects were detected very well.

• Korean Journal of Computational Design and Engineering
• /
• v.16 no.4
• /
• pp.277-284
• /
• 2011

In tangible augmented reality (AR) environments for virtual prototyping, the user interacts with virtual products by manipulating tangible objects with his or her hands, but the user often encounter awkward situations in which his or her hands are occluded by augmented virtual objects, which reduces both immersion and ease of interaction. In this paper, we present how to resolve such hand region occlusion in order to enhance natural interaction and immersive visualization. In the AR environment considered, we use two types (product-type and pointer-type) of tangible objects for tangible user interaction with a virtual product of interest. Holding the tangible objects with his or her hands, the user can create input events by touching specified regions of the product-type tangible object with the pointer-type tangible object. We developed a method for resolving hand region occlusion frequently arising during such user interaction, It first detect hand region in a real image and refines the rendered image of the virtual object by subtracting the hand region from the rendered image, Then, it superimposes the refined image onto the real image to obtain an image in which the occlusion is resolved. Incorporated into tangible AR interaction for virtual prototyping of handheld products such as cellular phones and MP3 players, the method has been found by a preliminary user study that it is not only useful to improve natural interaction and immersive visualization of virtual products, but also helpful for making the users experience the products' shapes and functions better.

• Journal of Drive and Control
• /
• v.19 no.3
• /
• pp.9-15
• /
• 2022

This paper proposes a method for constructing and verifying datasets used in deep learning technology, to prevent safety accidents in automated construction machinery or autonomous vehicles. Although open datasets for developing image recognition technologies are challenging to meet requirements desired by users, this study proposes the interface of virtual simulators to facilitate the creation of training datasets desired by users. The pixel-level training image dataset was verified by creating scenarios, including various road types and objects in a virtual environment. Detecting an object from an image may interfere with the accurate path determination due to occlusion areas covered by another object. Thus, we construct a database, for developing an occlusion area detection algorithm in a virtual environment. Additionally, we present the possibility of its use as a deep learning dataset to calculate a grid map, that enables path search considering occlusion areas. Custom datasets are built using the RDBMS system.

• The KIPS Transactions:PartB
• /
• v.13B no.1 s.104
• /
• pp.35-46
• /
• 2006

The dynamic change of background and moving objects is an important factor which causes the problem of occlusion in tracking moving objects. The tracking accuracy is also remarkably decreased in the presence of occlusion. We therefore propose an active fusion model which is robust against partial occlusions that are occurred by background and other objects. The active fusion model is consisted of contour-based md region-based snake. The former is a conventional snake model using contour features of a moving object and the latter is a regional snake model which considers region features inside its boundary. First, this model classifies total occlusion into contour and region occlusion. And then it adjusts the confidence of each model based on calculating the location and amount of occlusion, so it can overcome the problem of occlusion. Experimental results show that the proposed method can successfully track a moving object but the previous methods fail to track it under partial occlusion.

• The Journal of Engineering Research
• /
• v.2 no.1
• /
• pp.63-71
• /
• 1997

In this paper, we describe an approach that determines the identity of surfaces of an object with planar and curved surfaces from range data of the object in the scene. The proposed matching scheme presents that surface correspondence of an object is achieved by simple comparison of values for representing surfaces of the object with model in order to avoid unnecessary matching procedures. We use uniquely assigned Surface Representing Value(SRV) for representing surfaces of the object, which are sums of all weighted view-point independent features. And, the proposed method is simple, quite effective and insensitive to occlusion and noise in sensor data.