scholarly journals Hybrid features for object detection in RGB-D scenes

2021 ◽  
Vol 23 (2) ◽  
pp. 1073
Author(s):  
Sari Awwad ◽  
Bashar Igried ◽  
Mohammad Wedyan ◽  
Mohammad Alshira'H
Keyword(s):  
Object Detection ◽  
Sliding Window ◽  
Scale Invariant ◽  
Hybrid Features ◽  
Fine Grained ◽  
A New Technique ◽  
Improving Accuracy ◽  
Window Approach ◽  
Scale Invariant Feature ◽  
Local Depth

<div>Object detection is considered a hot research topic in applications of artificial intel-ligence and computer vision. Historically, object detection was widely used in var-ious fields like surveillance, fine-grained activities and robotics. All studies focus on improving accuracy for object detection using images, whether indoor or outdoor scenes. Therefore, this paper took a shot by improving the doable features extraction and proposing crossed sliding window approach using exiting classifiers for object de-tection. In this paper, the contribution includes two parts: First, improving local depth pattern feature along side SIFT and the second part explains a new technique presented by proposing crossed sliding window approach using two different types of images (colored and depth). Two types of features local depth patterns for detection (LDPD) and scale-invariant feature transform (SIFT) were merged as one feature vector. The RGB-D object dataset has been used and it consists of 300 different objects and in-cludes thousands of scenes. The proposed approach achieved high results comparing to other features or separated features that are used in this paper. All experiments and comparatives were applied on the same dataset for the same objective. Experimental results report a high accuracy in terms of detection rate, recall, precision and F1 scorein RGB-D scenes.</div>

Passive Copy- Move Forgery Detection Using Speed-Up Robust Features, Histogram Oriented Gradients and Scale Invariant Feature Transform

2015 ◽  
Vol 4 (3) ◽  
pp. 70-89
Author(s):  
Ramesh Chand Pandey ◽  
Sanjay Kumar Singh ◽  
K K Shukla
Keyword(s):  
Image Features ◽  
Forgery Detection ◽  
Scale Invariant ◽  
Hybrid Features ◽  
Single Region ◽  
Multiple Regions ◽  
Speed Up ◽  
Common Technique ◽  
Copy Move Forgery Detection ◽  
Scale Invariant Feature

Copy-Move is one of the most common technique for digital image tampering or forgery. Copy-Move in an image might be done to duplicate something or to hide an undesirable region. In some cases where these images are used for important purposes such as evidence in court of law, it is important to verify their authenticity. In this paper the authors propose a novel method to detect single region Copy-Move Forgery Detection (CMFD) using Speed-Up Robust Features (SURF), Histogram Oriented Gradient (HOG), Scale Invariant Features Transform (SIFT), and hybrid features such as SURF-HOG and SIFT-HOG. SIFT and SURF image features are immune to various transformations like rotation, scaling, translation, so SIFT and SURF image features help in detecting Copy-Move regions more accurately in compared to other image features. Further the authors have detected multiple regions COPY-MOVE forgery using SURF and SIFT image features. Experimental results demonstrate commendable performance of proposed methods.

Automatic Reel Editing in Chip on Film Quality Control With Computer Vision

2021 ◽  
Vol 11 (1) ◽  
pp. 1-14
Author(s):  
Shing Hwang Doong
Keyword(s):  
Quality Control ◽  
Integrated Circuits ◽  
Object Detection ◽  
Object Tracking ◽  
Template Matching ◽  
Scale Invariant ◽  
Detection Techniques ◽  
Novel Method ◽  
Scale Invariant Feature ◽  
Chip On Film

Chip on film (COF) is a special packaging technology to pack integrated circuits in a flexible carrier tape. Chips packed with COF are primarily used in the display industry. Reel editing is a critical step in COF quality control to remove sections of congregating NG (not good) chips from a reel. Today, COF manufactures hire workers to count consecutive NG chips in a rolling reel with naked eyes. When the count is greater than a preset number, the corresponding section is removed. A novel method using object detection and object tracking is proposed to solve this problem. Object detection techniques including convolutional neural network (CNN), template matching (TM), and scale invariant feature transform (SIFT) were used to detect NG marks, and object tracking was used to track them with IDs so that congregating NG chips could be counted reliably. Using simulation videos similar to worksite scenes, experiments show that both CNN and TM detectors could solve the reel editing problem, while SIFT detectors failed. Furthermore, TM is better than CNN by yielding a real time solution.

Sign in / Sign up

Export Citation Format

Share Document