João Marcos Silva, Romuere R. V. Silva, R. Veras, K. Aires, L. B. Britto Neto
Facial expression recognition systems can help a visually impaired person to identify the emotions of the person with whom she interacts, assisting in her non-verbal communication. Among the various researches carried out in recent years on recognition of facial expressions, the best results obtained come from methods that use deep learning, mainly with the use of convolutional neural networks. This work presents a literature review on the problem of recognition of facial expressions, through the use of convolutional neural networks and proposes two approaches in which the first one uses pre-trained CNN models together with the Linear SVM classifier that, applied to the bases CK+ and JAFFE data, obtained maximum accuracy of 89.6% and 95.7%, respectively. And in the second approach, a CNN model built from scratch is used with the CK+ and FER2013 databases, which obtained accuracy rates of 85% and 65.8%, respectively.
{"title":"Facial Expression Recognition to Aid Visually Impaired People","authors":"João Marcos Silva, Romuere R. V. Silva, R. Veras, K. Aires, L. B. Britto Neto","doi":"10.5753/wvc.2021.18888","DOIUrl":"https://doi.org/10.5753/wvc.2021.18888","url":null,"abstract":"Facial expression recognition systems can help a visually impaired person to identify the emotions of the person with whom she interacts, assisting in her non-verbal communication. Among the various researches carried out in recent years on recognition of facial expressions, the best results obtained come from methods that use deep learning, mainly with the use of convolutional neural networks. This work presents a literature review on the problem of recognition of facial expressions, through the use of convolutional neural networks and proposes two approaches in which the first one uses pre-trained CNN models together with the Linear SVM classifier that, applied to the bases CK+ and JAFFE data, obtained maximum accuracy of 89.6% and 95.7%, respectively. And in the second approach, a CNN model built from scratch is used with the CK+ and FER2013 databases, which obtained accuracy rates of 85% and 65.8%, respectively.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123735649","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jhonatan Souza, Claudemir Casa, André Roberto Ortoncelli
This work presents an analysis of the efficiency and effectiveness of a Video-Based Pain Monitoring System running on a Raspberry selected because it is a cheap device that can be easily carried around. The objective of the evaluated system is to allow the assessment of pain based on two characteristics: Heart Rate (HR) and facial expressions detected through the Facial Action Coding System (FACS). To measure HR an Eulerian Video Magnification (EVM) based method was implemented. EVM is one of the main current approaches to measure HR by Remote PhotoPlethysmoGraphy. FACS was used to detect pain intensity with the Prkachin and Solomon Pain Intensity (PSPI) equation which is one of the most used approaches to detect pain intensity based on facial features. To identify the PSPI value we trained a Regression Neural Network (RNN) with the UNBC-McMaster database. The experimental results demonstrate the strengths and limitations of the evaluated system.
{"title":"Analysis of a Video-Based Pain Monitoring System in Raspberry Pi","authors":"Jhonatan Souza, Claudemir Casa, André Roberto Ortoncelli","doi":"10.5753/wvc.2021.18913","DOIUrl":"https://doi.org/10.5753/wvc.2021.18913","url":null,"abstract":"This work presents an analysis of the efficiency and effectiveness of a Video-Based Pain Monitoring System running on a Raspberry selected because it is a cheap device that can be easily carried around. The objective of the evaluated system is to allow the assessment of pain based on two characteristics: Heart Rate (HR) and facial expressions detected through the Facial Action Coding System (FACS). To measure HR an Eulerian Video Magnification (EVM) based method was implemented. EVM is one of the main current approaches to measure HR by Remote PhotoPlethysmoGraphy. FACS was used to detect pain intensity with the Prkachin and Solomon Pain Intensity (PSPI) equation which is one of the most used approaches to detect pain intensity based on facial features. To identify the PSPI value we trained a Regression Neural Network (RNN) with the UNBC-McMaster database. The experimental results demonstrate the strengths and limitations of the evaluated system.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"173 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121720385","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Brazil is the largest coffee producer in the world, and then there are many challenges to maintain the high quality and purity of the beans. Thus, it is important to study coffee plants, and help agronomists to detect diseases, such as rust, with resources of computer science. In this work, it is described experiments using image segmentation algorithm JSEG, which is capable to segment images in multi-scale. Using a coffee tree image database RoCoLe (Robusta Coffee Leaf Images), the JSEG algorithm is used to segment these images in four scales. It is selected typical segments in each scale and they are grouped using similarity of normalized color histograms. In this way the several scales segmentations are compared. It is concluded that the segments in scales 1 and 2, in which the colors are more homogeneous then in scales 3 and 4, are adequate to use as training samples for the detection of rust diseases.
{"title":"Coffee plant image segmentation and disease detection using JSEG algorithm","authors":"Jeferson de Souza Dias, J. H. Saito","doi":"10.5753/wvc.2021.18887","DOIUrl":"https://doi.org/10.5753/wvc.2021.18887","url":null,"abstract":"Brazil is the largest coffee producer in the world, and then there are many challenges to maintain the high quality and purity of the beans. Thus, it is important to study coffee plants, and help agronomists to detect diseases, such as rust, with resources of computer science. In this work, it is described experiments using image segmentation algorithm JSEG, which is capable to segment images in multi-scale. Using a coffee tree image database RoCoLe (Robusta Coffee Leaf Images), the JSEG algorithm is used to segment these images in four scales. It is selected typical segments in each scale and they are grouped using similarity of normalized color histograms. In this way the several scales segmentations are compared. It is concluded that the segments in scales 1 and 2, in which the colors are more homogeneous then in scales 3 and 4, are adequate to use as training samples for the detection of rust diseases.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114208765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
GANs (Generative Adversarial Networks) were proposed to generate realistic synthetic images. In this work, we will discuss the use of GANs as alternative reconstruction of different fingerprint images from the original ones. The samples result in the same person fingerprint but obtained with other textures. Thus, it is intended to contribute to improving the method to increase databases with new samples, incorporating textures, when the quantities are insufficient for any purpose. To verify the similarity of the synthesized images with the original ones, a convolutional Xception network and the RMSE metric are used. The results obtained with fingerprint images of 3 persons, 20 of each finger, and 4 different textures, showed the tradeoff between similarity, recognizability, and the number of epochs of the Pix2pix training.
{"title":"Pix2pix network for fingerprint texture image synthesis","authors":"Jader dos Santos Teles Cordeiro, J. H. Saito","doi":"10.5753/wvc.2021.18882","DOIUrl":"https://doi.org/10.5753/wvc.2021.18882","url":null,"abstract":"GANs (Generative Adversarial Networks) were proposed to generate realistic synthetic images. In this work, we will discuss the use of GANs as alternative reconstruction of different fingerprint images from the original ones. The samples result in the same person fingerprint but obtained with other textures. Thus, it is intended to contribute to improving the method to increase databases with new samples, incorporating textures, when the quantities are insufficient for any purpose. To verify the similarity of the synthesized images with the original ones, a convolutional Xception network and the RMSE metric are used. The results obtained with fingerprint images of 3 persons, 20 of each finger, and 4 different textures, showed the tradeoff between similarity, recognizability, and the number of epochs of the Pix2pix training.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126646909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Adair da Silva Oliveira Junior, M. Pache, Fábio Prestes Cesar Rezende, D. Sant’Ana, V. Weber, Gilberto Astolfi, F. Weber, G. Menezes, Gabriel Kirsten Menezes, Pedro Lucas França Albuquerque, Celso Soares Costa, Vanir Garcia, Eduardo Quirino Arguelho de Queiroz, João Victor Araújo Rozales, M. Ferreira, M. Naka, H. Pistori
The objective of this paper is to investigate which combination of parameters for the fingerling counting software results in the smallest Mean Absolute Error (MAE) and smallest Root Mean Squared Error (RMSE). For this, an image dataset called FISHCV155V was created and separated into training and test sets, where different combinations of parameters for the software were tested. From the obtained results were extracted individual performance metrics for each combination of parameters, such as MAE, Mean Square Error (MSE) and RMSE. Video frames were analysed comparing the parameter combination that obtained the best and worst results, in order to investigate the influence of such parameters in the performance of the software. From such results, it was concluded that the best combination reached 5.99 MAE and 9.96 RMSE.
{"title":"An Investigation of Parameter Optimization in Fingerling Counting Problems","authors":"Adair da Silva Oliveira Junior, M. Pache, Fábio Prestes Cesar Rezende, D. Sant’Ana, V. Weber, Gilberto Astolfi, F. Weber, G. Menezes, Gabriel Kirsten Menezes, Pedro Lucas França Albuquerque, Celso Soares Costa, Vanir Garcia, Eduardo Quirino Arguelho de Queiroz, João Victor Araújo Rozales, M. Ferreira, M. Naka, H. Pistori","doi":"10.5753/wvc.2021.18881","DOIUrl":"https://doi.org/10.5753/wvc.2021.18881","url":null,"abstract":"The objective of this paper is to investigate which combination of parameters for the fingerling counting software results in the smallest Mean Absolute Error (MAE) and smallest Root Mean Squared Error (RMSE). For this, an image dataset called FISHCV155V was created and separated into training and test sets, where different combinations of parameters for the software were tested. From the obtained results were extracted individual performance metrics for each combination of parameters, such as MAE, Mean Square Error (MSE) and RMSE. Video frames were analysed comparing the parameter combination that obtained the best and worst results, in order to investigate the influence of such parameters in the performance of the software. From such results, it was concluded that the best combination reached 5.99 MAE and 9.96 RMSE.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"343 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123119254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Letícia Leite Caetano, Jocival Dantas Dias Júnior, A. Backes, H. Ferraz, M. Escarpinati
The growing use of technologies in favor of Precision Agriculture enables the application of different strategies in a crop and seeks to increase production, reduce costs and reduce damage to the environment. To keep up with the need to increase productivity and still reduce costs with farming as much as possible, the approach of applying inputs in a targeted manner based on the classification of regions is increasingly used, as are the results obtained in [9]. In optimizing these results, some points were identified that could be improved in relation to the vector data of the generated Management Zones, such as overlapping between different zones, invalid geometries, and a very large amount of points, which add unnecessary complexity to the file. This work proposes an algorithm that aims to optimize these Management Zone results in a shapefile, and aims to correct invalid geometries, reduce the number of points that define the shapes of the zones, and the correction of overlapping regions so that zones with lesser vigor have priority. In addition, an adjustment of the spacing between the geometries is made while correcting the overlap between different zones. As a result, a new shapefile is created, composed only of valid geometries, fewer points, and no overlaps between different Management Zones. Specialists evaluated the results obtained and indicated them as adequate to solve the problem.
{"title":"Optimization of Management Zones Shape Files","authors":"Letícia Leite Caetano, Jocival Dantas Dias Júnior, A. Backes, H. Ferraz, M. Escarpinati","doi":"10.5753/wvc.2021.18905","DOIUrl":"https://doi.org/10.5753/wvc.2021.18905","url":null,"abstract":"The growing use of technologies in favor of Precision Agriculture enables the application of different strategies in a crop and seeks to increase production, reduce costs and reduce damage to the environment. To keep up with the need to increase productivity and still reduce costs with farming as much as possible, the approach of applying inputs in a targeted manner based on the classification of regions is increasingly used, as are the results obtained in [9]. In optimizing these results, some points were identified that could be improved in relation to the vector data of the generated Management Zones, such as overlapping between different zones, invalid geometries, and a very large amount of points, which add unnecessary complexity to the file. This work proposes an algorithm that aims to optimize these Management Zone results in a shapefile, and aims to correct invalid geometries, reduce the number of points that define the shapes of the zones, and the correction of overlapping regions so that zones with lesser vigor have priority. In addition, an adjustment of the spacing between the geometries is made while correcting the overlap between different zones. As a result, a new shapefile is created, composed only of valid geometries, fewer points, and no overlaps between different Management Zones. Specialists evaluated the results obtained and indicated them as adequate to solve the problem.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130091389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Éverton C. Acchetta, Lucas P. Laheras, Helmuth A. Risch, Vinicius L. O. P. Santos, P. S. Rodrigues
Investments in Augmented Reality (AR) have grown considerably in recent years. This advance is due to the increased use of AR in areas such as education, training, games and medicine. In addition, technological advances in hardware enable devices that, a few years ago, were unthinkable. A popular example is Microsoft Hololens 2, which allows the user to use their own hands as a means of interacting with an AR experience. However, a disadvantage from this device is its high cost due to several sensors. Thus, this project offers an AR architecture that uses only a monocular RGB camera as a sensor, allowing the user to interact with an AR experience using their hands to perform gestures similar to the Microsoft Hololens 2 architecture, where it is possible to handle a virtual object in the same way that a real object would be manipulated. The results obtained are promising, where the verification of the interaction of the hand with the virtual object worked in approximately 80% of the tests carried out, respecting the path defined by hand movement.
{"title":"Methodology and Implementation of an Architecture for Egocentric Manual Interactivity in Monocular Augmented Reality","authors":"Éverton C. Acchetta, Lucas P. Laheras, Helmuth A. Risch, Vinicius L. O. P. Santos, P. S. Rodrigues","doi":"10.5753/wvc.2021.18902","DOIUrl":"https://doi.org/10.5753/wvc.2021.18902","url":null,"abstract":"Investments in Augmented Reality (AR) have grown considerably in recent years. This advance is due to the increased use of AR in areas such as education, training, games and medicine. In addition, technological advances in hardware enable devices that, a few years ago, were unthinkable. A popular example is Microsoft Hololens 2, which allows the user to use their own hands as a means of interacting with an AR experience. However, a disadvantage from this device is its high cost due to several sensors. Thus, this project offers an AR architecture that uses only a monocular RGB camera as a sensor, allowing the user to interact with an AR experience using their hands to perform gestures similar to the Microsoft Hololens 2 architecture, where it is possible to handle a virtual object in the same way that a real object would be manipulated. The results obtained are promising, where the verification of the interaction of the hand with the virtual object worked in approximately 80% of the tests carried out, respecting the path defined by hand movement.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123056103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. Gomide, Elton Vieira Cunha, Guilherme Boechat Gomide
This paper presents the development of a computer vision system that automatically identifies and counts viable and inviable brewer's yeast, to improve the time and accuracy of results obtained compared to the manual expert counting method commonly performed in the brewing industry. The equipment used consists of a digital video camera coupled to an optical microscope, which transmits the captured images, in real time, to the computer. Two approaches were tested and implemented, one taking into account the morphology and color of yeasts, and the other using machine learning. Although there are programs that automatically count yeasts, this is the first application that makes use of convolutional neural network techniques with Yolo to identify yeasts, making the results more accurate and reliable compared to manual methods. Experiments were carried out to measure the performance and accuracy of the prototype, which are presented in this article.
{"title":"Automatic Yeast Detection and Counting Using Computer Vision Techniques","authors":"J. Gomide, Elton Vieira Cunha, Guilherme Boechat Gomide","doi":"10.5753/wvc.2021.18884","DOIUrl":"https://doi.org/10.5753/wvc.2021.18884","url":null,"abstract":"This paper presents the development of a computer vision system that automatically identifies and counts viable and inviable brewer's yeast, to improve the time and accuracy of results obtained compared to the manual expert counting method commonly performed in the brewing industry. The equipment used consists of a digital video camera coupled to an optical microscope, which transmits the captured images, in real time, to the computer. Two approaches were tested and implemented, one taking into account the morphology and color of yeasts, and the other using machine learning. Although there are programs that automatically count yeasts, this is the first application that makes use of convolutional neural network techniques with Yolo to identify yeasts, making the results more accurate and reliable compared to manual methods. Experiments were carried out to measure the performance and accuracy of the prototype, which are presented in this article.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116947163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
João Gondim, M. Maia, Ana Caroline Lopes Rocha, Felipe Argolo, Anderson Ara, A. Loch
Support Vector Machines are a set of machine learning models that have great performance in several tasks as well as on image classification and object recognition. However, the proper choice of model's hyperparameters has a great influence on the outcomes and the general capacity performance. In this paper, we explore some different traditional auto-tuning processes to estimate σ hyper-parameter for SVMs Gaussian kernel. These processes are common and also implemented on standard software of data science languages. The paper considers some different situations on smile detection. The results are composed by simulation study, two benchmark image applications and a real video data application.
{"title":"Support Vector Machines in Smile detection: A comparison of auto-tuning standard processes in Gaussian kernel","authors":"João Gondim, M. Maia, Ana Caroline Lopes Rocha, Felipe Argolo, Anderson Ara, A. Loch","doi":"10.5753/wvc.2021.18900","DOIUrl":"https://doi.org/10.5753/wvc.2021.18900","url":null,"abstract":"Support Vector Machines are a set of machine learning models that have great performance in several tasks as well as on image classification and object recognition. However, the proper choice of model's hyperparameters has a great influence on the outcomes and the general capacity performance. In this paper, we explore some different traditional auto-tuning processes to estimate σ hyper-parameter for SVMs Gaussian kernel. These processes are common and also implemented on standard software of data science languages. The paper considers some different situations on smile detection. The results are composed by simulation study, two benchmark image applications and a real video data application.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"392 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128340384","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The disease detection is vital to increase the productivity and quality of soybean cultivation and this detection is usually carried out in a laboratory, which is time consuming and costly. To overcome these issues, there is a growing demand for technologies that aim at a faster detection and classification of diseases. In this context, this work proposes the extraction of several patches from a leaf image and combining a convolutional neural network with a support vector machine, we present a complete model for the classification of soybean leaf diseases. In this approach, an image dataset with evidence of diseases commonly observed in soybean crops was analyzed and our experiments achieved precisions greater than 90%.
{"title":"Patch-Based Model for the Classification of Soybean Leaf Diseases","authors":"Gustavo Vigilato G. S., P. G. Cavalcanti","doi":"10.5753/wvc.2021.18899","DOIUrl":"https://doi.org/10.5753/wvc.2021.18899","url":null,"abstract":"The disease detection is vital to increase the productivity and quality of soybean cultivation and this detection is usually carried out in a laboratory, which is time consuming and costly. To overcome these issues, there is a growing demand for technologies that aim at a faster detection and classification of diseases. In this context, this work proposes the extraction of several patches from a leaf image and combining a convolutional neural network with a support vector machine, we present a complete model for the classification of soybean leaf diseases. In this approach, an image dataset with evidence of diseases commonly observed in soybean crops was analyzed and our experiments achieved precisions greater than 90%.","PeriodicalId":311431,"journal":{"name":"Anais do XVII Workshop de Visão Computacional (WVC 2021)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124570966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：