scholarly journals A Multispectral Pest-Detection Algorithm for Precision Agriculture

2021 ◽  
Vol 12 (1) ◽  
pp. 46
Author(s):  
Syed Umar Rasheed ◽  
Wasif Muhammad ◽  
Irfan Qaiser ◽  
Muhammad Jehanzeb Irshad
Keyword(s):  
Deep Learning ◽  
Environmental Conditions ◽  
Precision Agriculture ◽  
Feature Fusion ◽  
Detection Algorithm ◽  
Type I ◽  
Multispectral Images ◽  
Pest Detection ◽  
The Mean ◽  
State Of Art

Invertebrates are abundant in horticulture and farming environments, and can be detrimental. Early pest detection for an integrated pest-management system with an integration of physical, biological, and prophylactic methods has huge potential for the better yield of crops. Computer vision techniques with multispectral images are used to detect and classify pests in dynamic environmental conditions, such as sunlight variations, partial occlusions, low contrast, etc. Various state-of-art, deep learning approaches have been proposed, but there are some major limitations to these methods. For example, labelled images are required to supervise the training of deep networks, which is tiresome work. Secondly, a huge in-situ database with variant environmental conditions is not available for deep learning, or is difficult to build for fretful bioaggressors. In this paper, we propose a machine-vision-based multispectral pest-detection algorithm, which does not require any kind of supervised network training. Multispectral images are used as input for the proposed pest-detection algorithm, and each image provides comprehensive information about different textural and morphological features, and visible information, i.e., size, shape, orientation, color, and wing patterns for each insect. Feature identification is performed by a SURF algorithm, and feature extraction is accomplished by least median of square regression (LMEDS). Feature fusion of RGB and NIR images onto the coordinates of Ultraviolet (UV) is performed after affine transformation. The mean identification errors of type I, II, and total mean error surpass the mean errors of the state-of-art methods. The type I, II, and total mean errors, with 6.672% UV weights, were emanated to 1.62, 40.27, and 3.26, respectively.

scholarly journals Identification of dental implants using deep learning—pilot study

2020 ◽  
Vol 6 (1) ◽  
Author(s):  
Toshihito Takahashi ◽  
Kazunori Nozaki ◽  
Tomoya Gonda ◽  
Tomoaki Mameno ◽  
Masahiro Wada ◽  
...  
Keyword(s):  
Deep Learning ◽  
Object Detection ◽  
Panoramic Radiograph ◽  
Detection Algorithm ◽  
Learning Performance ◽  
Identification System ◽  
Limited Data ◽  
Radiographic Images ◽  
The Mean ◽  
Implant System

Abstract Background In some cases, a dentist cannot solve the difficulties a patient has with an implant because the implant system is unknown. Therefore, there is a need for a system for identifying the implant system of a patient from limited data that does not depend on the dentist’s knowledge and experience. The purpose of this study was to identify dental implant systems using a deep learning method. Methods A dataset of 1282 panoramic radiograph images with implants were used for deep learning. An object detection algorithm (Yolov3) was used to identify the six implant systems by three manufactures. To implement the algorithm, TensorFlow and Keras deep-learning libraries were used. After training was complete, the true positive (TP) ratio and average precision (AP) of each implant system as well as the mean AP (mAP), and mean intersection over union (mIoU) were calculated to evaluate the performance of the model. Results The number of each implant system varied from 240 to 1919. The TP ratio and AP of each implant system varied from 0.50 to 0.82 and from 0.51 to 0.85, respectively. The mAP and mIoU of this model were 0.71 and 0.72, respectively. Conclusions The results of this study suggest that implants can be identified from panoramic radiographic images using deep learning-based object detection. This identification system could help dentists as well as patients suffering from implant problems. However, more images of other implant systems will be necessary to increase the learning performance to apply this system in clinical practice.

scholarly journals Lightweight Fruit-Detection Algorithm for Edge Computing Applications

2021 ◽  
Vol 12 ◽  
Author(s):  
Wenli Zhang ◽  
Yuxin Liu ◽  
Kaizhen Chen ◽  
Huibin Li ◽  
Yulin Duan ◽  
...  
Keyword(s):  
Deep Learning ◽  
Real Time ◽  
Sampling Method ◽  
Feature Fusion ◽  
Detection Algorithm ◽  
Detection Accuracy ◽  
Model Structure ◽  
Excellent Performance ◽  
Backbone Network ◽  
Detection Technology

In recent years, deep-learning-based fruit-detection technology has exhibited excellent performance in modern horticulture research. However, deploying deep learning algorithms in real-time field applications is still challenging, owing to the relatively low image processing capability of edge devices. Such limitations are becoming a new bottleneck and hindering the utilization of AI algorithms in modern horticulture. In this paper, we propose a lightweight fruit-detection algorithm, specifically designed for edge devices. The algorithm is based on Light-CSPNet as the backbone network, an improved feature-extraction module, a down-sampling method, and a feature-fusion module, and it ensures real-time detection on edge devices while maintaining the fruit-detection accuracy. The proposed algorithm was tested on three edge devices: NVIDIA Jetson Xavier NX, NVIDIA Jetson TX2, and NVIDIA Jetson NANO. The experimental results show that the average detection precision of the proposed algorithm for orange, tomato, and apple datasets are 0.93, 0.847, and 0.850, respectively. Deploying the algorithm, the detection speed of NVIDIA Jetson Xavier NX reaches 21.3, 24.8, and 22.2 FPS, while that of NVIDIA Jetson TX2 reaches 13.9, 14.1, and 14.5 FPS and that of NVIDIA Jetson NANO reaches 6.3, 5.0, and 8.5 FPS for the three datasets. Additionally, the proposed algorithm provides a component add/remove function to flexibly adjust the model structure, considering the trade-off between the detection accuracy and speed in practical usage.

scholarly journals Physical characteristics and spillage detection Using multi-feature fusion

2021 ◽  
Vol 2083 (2) ◽  
pp. 022041
Author(s):  
Caiyu Liu ◽  
Zuofeng Zhou ◽  
Qingquan Wu
Keyword(s):  
Deep Learning ◽  
Clustering Algorithm ◽  
Feature Fusion ◽  
Detection Algorithm ◽  
Detection Accuracy ◽  
Complex Data ◽  
Learning Approaches ◽  
Road Maintenance ◽  
Data Set ◽  
First Time

Abstract As an important part of road maintenance, the detection of road sprinkles has attracted extensive attention from scholars. However, after years of research, there are still some problems in the detection of road sprinkles. First of all, the detection accuracy of traditional detection algorithm is deficient. Second, deep learning approaches have great limitations for there are various kinds of sprinkles which makes it difficult to build a data set. In view of the above problems, this paper proposes a road sprinkling detection method based on multi-feature fusion. The characteristics of color, gradient, luminance and neighborhood information were considered in our method. Compared with other traditional methods, our method has higher detection accuracy. In addition, compared with deep learning-based methods, our approach doesn’t involve creating a complex data set and reduces costs. The main contributions of this paper are as follows: I. For the first time, the density clustering algorithm is combined with the detection of sprinkles, which provides a new idea for this field. II. The use of multi-feature fusion improves the accuracy and robustness of the traditional method which makes the algorithm usable in many real-world scenarios.

A fast workpiece detection method based on multi-feature fused SSD

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Guoyuan Shi ◽  
Yingjie Zhang ◽  
Manni Zeng
Keyword(s):  
Deep Learning ◽  
Feature Fusion ◽  
Learning Algorithms ◽  
Detection Algorithm ◽  
Detection Accuracy ◽  
Data Sets ◽  
Single Shot ◽  
Content Type ◽  
Sorting System ◽  
Study Designs

Purpose Workpiece sorting is a key link in industrial production lines. The vision-based workpiece sorting system is non-contact and widely applicable. The detection and recognition of workpieces are the key technologies of the workpiece sorting system. To introduce deep learning algorithms into workpiece detection and improve detection accuracy, this paper aims to propose a workpiece detection algorithm based on the single-shot multi-box detector (SSD). Design/methodology/approach Propose a multi-feature fused SSD network for fast workpiece detection. First, the multi-view CAD rendering images of the workpiece are used as deep learning data sets. Second, the visual geometry group network was trained for workpiece recognition to identify the category of the workpiece. Third, this study designs a multi-level feature fusion method to improve the detection accuracy of SSD (especially for small objects); specifically, a feature fusion module is added, which uses “element-wise sum” and “concatenation operation” to combine the information of shallow features and deep features. Findings Experimental results show that the actual workpiece detection accuracy of the method can reach 96% and the speed can reach 41 frames per second. Compared with the original SSD, the method improves the accuracy by 7% and improves the detection performance of small objects. Originality/value This paper innovatively introduces the SSD detection algorithm into workpiece detection in industrial scenarios and improves it. A feature fusion module has been added to combine the information of shallow features and deep features. The multi-feature fused SSD network proves the feasibility and practicality of introducing deep learning algorithms into workpiece sorting.

scholarly journals Research on Object Detection Algorithm Based on Multilayer Information Fusion

2020 ◽  
Vol 2020 ◽  
pp. 1-13
Author(s):  
Bao-Yuan Chen ◽  
Yu-Kun Shen ◽  
Kun Sun
Keyword(s):  
Feature Extraction ◽  
Object Detection ◽  
Feature Fusion ◽  
Basic Feature ◽  
Detection Algorithm ◽  
Mean Average Precision ◽  
Detection Accuracy ◽  
Average Precision ◽  
Position Information ◽  
The Mean

At present, object detectors based on convolution neural networks generally rely on the last layer of features extracted by the feature extraction network. In the process of continuous convolution and pooling of deep features, the position information cannot be completely transferred backward. This paper proposes a multiscale feature reuse detection model, which includes the basic feature extraction network DenseNet, feature fusion network, multiscale anchor region proposal network, and classification and regression network. The fusion of high-dimensional features and low-dimensional features not only strengthens the model's sensitivity to objects of different sizes but also strengthens the transmission of information, so that the feature map has rich deep semantic information and shallow location information at the same time, which significantly improves the robustness and detection accuracy of the model. The algorithm is trained and tested in Pascal VOC2007 dataset. The experimental results show that the mean average precision of the objects in the dataset is 73.87%. At the same time, compared with the mainstream faster RCNN and SSD detection models, the mean average precision of object detection algorithm based on DenseNet is improved by 5.63% and 3.86%, respectively.

Comparative Study of effectiveness of mobilization with movement (MWM) and End range mobilization (ERM) techniques in frozen shoulder

2018 ◽  
Vol 4 (4) ◽  
pp. 519-522
Author(s):  
Jeyakumar S ◽  
Jagatheesan Alagesan ◽  
T.S. Muthukumar
Keyword(s):  
Range Of Motion ◽  
Frozen Shoulder ◽  
Type I ◽  
Mean Values ◽  
Functional Activities ◽  
Group A ◽  
The Mean ◽  
Group B

Background: Frozen shoulder is disorder of the connective tissue that limits the normal Range of motion of the shoulder in diabetes, frozen shoulder is thought to be caused by changes to the collagen in the shoulder joint as a result of long term Hypoglycemia. Mobilization is a therapeutic movement of the joint. The goal is to restore normal joint motion and rhythm. The use of mobilization with movement for peripheral joints was developed by mulligan. This technique combines a sustained application of manual technique “gliding” force to the joint with concurrent physiologic motion of joint, either actively or passively. This study aims to find out the effects of mobilization with movement and end range mobilization in frozen shoulder in Type I diabetics. Materials and Methods: 30 subjects both male and female, suffering with shoulder pain and clinically diagnosed with frozen shoulder was recruited for the study and divided into two groups with 15 patients each based on convenient sampling method. Group A patients received mobilization with movement and Group B patients received end range mobilization for three weeks. The outcome measurements were SPADI, Functional hand to back scale, abduction range of motion using goniometer and VAS. Results: The mean values of all parameters showed significant differences in group A as compared to group B in terms of decreased pain, increased abduction range and other outcome measures. Conclusion: Based on the results it has been concluded that treating the type 1 diabetic patient with frozen shoulder, mobilization with movement exercise shows better results than end range mobilization in reducing pain and increase functional activities and mobility in frozen shoulder.

Radiological and clinical associations with scoliosis outcomes after posterior fossa decompression in patients with Chiari malformation and syrinx from the Park-Reeves Syringomyelia Research Consortium

2020 ◽  
Vol 26 (1) ◽  
pp. 53-59 ◽  
Author(s):  
Jennifer M. Strahle ◽  
Rukayat Taiwo ◽  
Christine Averill ◽  
James Torner ◽  
Jordan I. Gewirtz ◽  
...  
Keyword(s):  
Posterior Fossa ◽  
Chiari Malformation ◽  
Craniocervical Junction ◽  
Curve Progression ◽  
Posterior Fossa Decompression ◽  
Type I ◽  
Younger Age ◽  
Curve Magnitude ◽  
The Mean

OBJECTIVEIn patients with Chiari malformation type I (CM-I) and a syrinx who also have scoliosis, clinical and radiological predictors of curve regression after posterior fossa decompression are not well known. Prior reports indicate that age younger than 10 years and a curve magnitude < 35° are favorable predictors of curve regression following surgery. The aim of this study was to determine baseline radiological factors, including craniocervical junction alignment, that might predict curve stability or improvement after posterior fossa decompression.METHODSA large multicenter retrospective and prospective registry of pediatric patients with CM-I (tonsils ≥ 5 mm below the foramen magnum) and a syrinx (≥ 3 mm in width) was reviewed for clinical and radiological characteristics of CM-I, syrinx, and scoliosis (coronal curve ≥ 10°) in patients who underwent posterior fossa decompression and who also had follow-up imaging.RESULTSOf 825 patients with CM-I and a syrinx, 251 (30.4%) were noted to have scoliosis present at the time of diagnosis. Forty-one (16.3%) of these patients underwent posterior fossa decompression and had follow-up imaging to assess for scoliosis. Twenty-three patients (56%) were female, the mean age at time of CM-I decompression was 10.0 years, and the mean follow-up duration was 1.3 years. Nine patients (22%) had stable curves, 16 (39%) showed improvement (> 5°), and 16 (39%) displayed curve progression (> 5°) during the follow-up period. Younger age at the time of decompression was associated with improvement in curve magnitude; for those with curves of ≤ 35°, 17% of patients younger than 10 years of age had curve progression compared with 64% of those 10 years of age or older (p = 0.008). There was no difference by age for those with curves > 35°. Tonsil position, baseline syrinx dimensions, and change in syrinx size were not associated with the change in curve magnitude. There was no difference in progression after surgery in patients who were also treated with a brace compared to those who were not treated with a brace for scoliosis.CONCLUSIONSIn this cohort of patients with CM-I, a syrinx, and scoliosis, younger age at the time of decompression was associated with improvement in curve magnitude following surgery, especially in patients younger than 10 years of age with curves of ≤ 35°. Baseline tonsil position, syrinx dimensions, frontooccipital horn ratio, and craniocervical junction morphology were not associated with changes in curve magnitude after surgery.

COLONY OF THE FUNGUS Aspergillus oryzae AND SELF-AFFINE FRACTAL GEOMETRY OF GROWTH FRONTS

Fractals ◽  
1993 ◽  
Vol 01 (01) ◽  
pp. 11-19 ◽  
Author(s):  
SHU MATSUURA ◽  
SASUKE MIYAZIMA
Keyword(s):  
Aspergillus Oryzae ◽  
Environmental Conditions ◽  
Fractal Geometry ◽  
Nutrient Concentration ◽  
Incubation Temperature ◽  
Characteristic Exponent ◽  
Growth Front ◽  
Favorable Condition ◽  
The Self ◽  
The Mean

A variety of colony shapes of the fungus Aspergillus oryzae under varying environmental conditions such as the nutrient concentration, medium stiffness and incubation temperature are obtained, ranging from a homogeneous Eden-like to a ramified DLA-like pattern. The roughness σ(l, h) of the growth front of the band-shaped colony, where h is the mean front height within l of the horizontal range, satisfies the self-affine fractal relation under favorable environmental conditions. In the most favorable condition of our experiments, its characteristic exponent is found to be a little larger than that of the 2-dimensional Eden model.

Sign in / Sign up

Export Citation Format

Share Document