scholarly journals Depth Image Processing and Operator Imitation Using a Custom Made Semi Humanoid

2012 ◽  
Vol 1 (1) ◽  
pp. 31-37
Author(s):  
Ghanshyam Bhutra
Keyword(s):  
Image Processing ◽  
Depth Image ◽  
Custom Made

scholarly journals Depth image correction for intel realsense depth camera

2020 ◽  
Vol 19 (2) ◽  
pp. 1021
Author(s):  
Hyun Jun Park ◽  
Kwang Baek Kim
Keyword(s):  
Image Processing ◽  
Infrared Radiation ◽  
Infrared Camera ◽  
High Accuracy ◽  
Depth Image ◽  
Depth Information ◽  
Depth Camera ◽  
Image Correction ◽  
Image Processing Techniques ◽  
Processing Techniques

<p><span>Intel RealSense depth camera provides depth image using infrared projector and infrared camera. Using infrared radiation makes it possible to measure the depth with high accuracy, but the shadow of infrared radiation makes depth unmeasured regions. Intel RealSense SDK provides a postprocessing algorithm to correct it. However, this algorithm is not enough to be used and needs to be improved. Therefore, we propose a method to correct the depth image using image processing techniques. The proposed method corrects the depth using the adjacent depth information. Experimental results showed that the proposed method corrects the depth image more accurately than the Intel RealSense SDK.</span></p>

scholarly journals Robotic Plug-in Combined Charging System with Improved Robustness

Proceedings ◽  
2019 ◽  
Vol 42 (1) ◽  
pp. 11
Author(s):  
Josef Cernohorsky ◽  
Pavel Jandura
Keyword(s):  
Image Processing ◽  
Inertial Sensor ◽  
Mobile Platform ◽  
Depth Image ◽  
Charging System ◽  
Internal Parameters ◽  
Six Dof ◽  
Under Construction ◽  
Conventional Solution

This paper describes the development of an algorithm robotic plug-in of a charging system for mobile platform. In the first chapter, there is a short overview of possibilities of automatic plug-in system, including proprietary industrial solution. In the main part, there is a description of the system based on UR robot with build-in force torque sensors and Intel RealSense Camera. This camera combines IR depth lens with regular RGB camera and six DOF inertial sensor, which is used in our application too. The conventional solution of this problem is usually based on RGB image processing in various state of the art, from simple pattern matching, neural network, or genetic algorithm to complex AI solution. The quality of the solution mostly depends on robustness of image processing. In our cases, we use simple sensor fusion. Thanks to multiple information and constrain of values, we can assume, if the algorithm is proceeding successfully or not. The system uses the internal parameters of the robotic arm, e.g., end-effector position and orientation and force-torque information in tool center point. The next information is RGB camera image and camera depth image, and the inertial unit build in camera. The other important information is the location of the vehicle inlet on the mobile platform, where the shape of mobile platform is considered as a constrain for image processing. The system is validated only on a physical model with CCS type 2 plug and vehicle inlet, because the mobile platform is under construction by another team.

scholarly journals Obstacle detection using Microsoft Kinect

2021 ◽  
Author(s):  
Niclas Zeller
Keyword(s):  
Image Processing ◽  
Obstacle Detection ◽  
Microsoft Kinect ◽  
Depth Image ◽  
Normal Vector ◽  
Blind Person ◽  
Camera System ◽  
Real Time System ◽  
Plane Fitting ◽  
Kinect Camera

This thesis presents the development of image processing algorithms based on a Microsoft Kinect camera system. The algorithms developed during this thesis are applied on the depth image received from Kinect and are supposed to model a three dimensional object based representation of the recorded scene. The motivation behind this thesis is to develop a system which assists visually impaired people by navigating through unknown environments. The developed system is able to detect obstacles in the recorded scene and to warn about these obstacles. Since the goal of this thesis was not to develop a complete real time system but to invent reliable algorithms solving this task, the algorithms were developed in MATLAB. Additionally a control software was developed by which depth as well as color images can be received from Kinect. The developed algorithms are a combination of already known plane fitting algorithms and novel approaches. The algorithms perform a plane segmentation of the 3D point cloud and model objects out of the received segments. Each obstacle is defined by a cuboid box and thus can be illustrated easily to the blind person. For plane segmentation different approaches were compared to each other to find the most suitable approach. The first algorithm analyzed in this thesis is a normal vector based plane fitting algorithm. This algorithm supplies very accurate results but also has a high computation effort. The second approach, which was finally implemented, is a gradient based 2D image segmentation combined with a RANSAC plane segmentation (6) in a 3D points cloud. This approach has the advantage to find very small edges within the scene but also builds planes based on global constrains. Beside the development of the algorithm results of the image processing, which are really promising, are presented. Thus the algorithm is worth to be improved by further development. The developed algorithm is able to detect very small but significant obstacles but on the other hand does not represent the scene too detailed such that the result can be illustrated accurately to a blind person.

Measuring the Cap Diameter of White Button Mushrooms ( Agaricus bisporus ) by Using Depth Image Processing

2021 ◽  
Vol 37 (4) ◽  
pp. 623-633
Author(s):  
Jiangtao Ji ◽  
Jingwei Sun ◽  
Xin Jin ◽  
Hao Ma ◽  
Xuefeng Zhu
Keyword(s):  
Image Processing ◽  
Measurement Error ◽  
Agaricus Bisporus ◽  
Detection Rate ◽  
Depth Image ◽  
Diameter Measurement ◽  
Selective Harvesting ◽  
Background Segmentation ◽  
Button Mushrooms ◽  
White Button Mushrooms

Highlights A new background segmentation algorithm for depth image was developed. Cap diameter of white button mushroom was measured automatically. The average of diameter measurement error was 4.94%. This work can provide online decision support for selectively harvesting of Agaricus bisporus . Abstract. With the increase in the production and yield of white button mushrooms (Agaricus bisporus), efficient harvesting has become a challenge. Automatic selective harvesting has gradually become a solution. The diameter of the mushroom cap is an essential indicator of the harvesting standard. To provide guidance for selective harvesting, this article presents a method for target detection and measuring the diameter of mushroom caps by using depth image processing. According to the three-dimensional structure characteristics of the mushroom, a novel method is proposed to segment it from the compost it grows on. In this method, compost is regarded as the floor of the sea and mushrooms as standing islands. With the rise of sea level, the compost is gradually submerged, and the target of Agaricus bisporus is stable. These features were used to realize the background segmentation. After background segmentation, the pixel coordinates of the contour points of the mushroom caps are transformed into world coordinates, and the cap diameter is measured by Hough transform. In total, 380 mushrooms depicted in 25 depth images were used to test the developed algorithms. The results showed that 92.37% of the mushrooms were correctly detected. The missed detection rate was less than 8%, and the false detection rate was 1.96%. The average diameter measurement error was 4.94%, and the average process time to measure a single mushroom was approximately 0.50 s. The method proposed in this article can provide online decision support for automatic selective harvesting of Agaricus bisporus, which can improve the quality and efficiency of its production. Keywords: Background segmentation, Computer vision, Diameter measurement, Edible fungus, Hough transform.

scholarly journals Obstacle detection using Microsoft Kinect

2021 ◽  
Author(s):  
Niclas Zeller
Keyword(s):  
Image Processing ◽  
Obstacle Detection ◽  
Microsoft Kinect ◽  
Depth Image ◽  
Normal Vector ◽  
Blind Person ◽  
Camera System ◽  
Real Time System ◽  
Plane Fitting ◽  
Kinect Camera

This thesis presents the development of image processing algorithms based on a Microsoft Kinect camera system. The algorithms developed during this thesis are applied on the depth image received from Kinect and are supposed to model a three dimensional object based representation of the recorded scene. The motivation behind this thesis is to develop a system which assists visually impaired people by navigating through unknown environments. The developed system is able to detect obstacles in the recorded scene and to warn about these obstacles. Since the goal of this thesis was not to develop a complete real time system but to invent reliable algorithms solving this task, the algorithms were developed in MATLAB. Additionally a control software was developed by which depth as well as color images can be received from Kinect. The developed algorithms are a combination of already known plane fitting algorithms and novel approaches. The algorithms perform a plane segmentation of the 3D point cloud and model objects out of the received segments. Each obstacle is defined by a cuboid box and thus can be illustrated easily to the blind person. For plane segmentation different approaches were compared to each other to find the most suitable approach. The first algorithm analyzed in this thesis is a normal vector based plane fitting algorithm. This algorithm supplies very accurate results but also has a high computation effort. The second approach, which was finally implemented, is a gradient based 2D image segmentation combined with a RANSAC plane segmentation (6) in a 3D points cloud. This approach has the advantage to find very small edges within the scene but also builds planes based on global constrains. Beside the development of the algorithm results of the image processing, which are really promising, are presented. Thus the algorithm is worth to be improved by further development. The developed algorithm is able to detect very small but significant obstacles but on the other hand does not represent the scene too detailed such that the result can be illustrated accurately to a blind person.

scholarly journals Segmentation of Tomato Growing Truss a Depth Image Conversion Model Based on CycleGAN

2021 ◽  
Author(s):  
Dae-Hyun Jung ◽  
Cheoul Young Kim ◽  
Taek Sung Lee ◽  
Soo Hyun Park
Keyword(s):  
Image Processing ◽  
Evaluation Method ◽  
False Negative ◽  
Depth Image ◽  
Tomato Plants ◽  
Conversion Model ◽  
Depth Images ◽  
Image Conversion ◽  
Straight Line ◽  
Artificial Intelligence Technology

Abstract Background: The truss on tomato plants is a group or cluster of smaller stems where flowers and fruit develop, while a growing truss is the most extended part of the stem. Because the state of the growing truss reacts sensitively to the surrounding environment, it is essential to control the growth in the early stages. With the recent development of IT and artificial intelligence technology in agriculture, a previous study developed a real-time acquisition and evaluation method for images using robots. Further, we used image processing to locate the growing truss and flowering rooms to extract growth information such as the height of the flower room and hard crab. Among the different vision algorithms, the CycleGAN algorithm was used to generate and transform unpaired images using generatively learning images. In this study, we developed a robot-based system for simultaneously acquiring RGB and depth images of the tomato growing truss and flower room groups.Results: The segmentation performance for approximately 35 samples was compared through the false negative (FN) and false positive (FP) indicators. For the depth camera image, we obtained FN as 17.55±3.01% and FP as 17.76±3.55%. Similarly, for CycleGAN, we obtained FN as approximately 19.24±1.45% and FP as 18.24±1.54%. As a result of image processing through depth image, IoU was 63.56 ± 8.44%, and when segmentation was performed through CycelGAN, IoU was 69.25 ± 4.42%, indicating that CycleGAN is advantageous in extracting the desired growing truss. Conclusions: The scannability was confirmed when the image scanning robot drove in a straight line through the plantation in the tomato greenhouse, which confirmed the on-site possibility of the image extraction technique using CycleGAN. In the future, the proposed approach is expected to be used in vision technology to scan the tomato growth indicators in greenhouses using an unmanned robot platform.

scholarly journals Automatic recognition of lactating sow behaviors through depth image processing

2016 ◽  
Vol 125 ◽  
pp. 56-62 ◽  
Author(s):  
F. Lao ◽  
T. Brown-Brandl ◽  
J.P. Stinn ◽  
K. Liu ◽  
G. Teng ◽  
...  
Keyword(s):  
Image Processing ◽  
Automatic Recognition ◽  
Depth Image ◽  
Lactating Sow
Sign in / Sign up

Export Citation Format

Share Document