Apple Detection in Complex Scene Using the Improved YOLOv4 Model
TL;DR: The test results show that the EfficientNet-B0-Y OLOv4 model proposed in this paper has better detection performance than YOLOv3, YOLov4, and Faster R-CNN with ResNet, which are state-of-the-art apple detection model.
Abstract: To enable the apple picking robot to quickly and accurately detect apples under the complex background in orchards, we propose an improved You Only Look Once version 4 (YOLOv4) model and data augmentation methods. Firstly, the crawler technology is utilized to collect pertinent apple images from the Internet for labeling. For the problem of insufficient image data caused by the random occlusion between leaves, in addition to traditional data augmentation techniques, a leaf illustration data augmentation method is proposed in this paper to accomplish data augmentation. Secondly, due to the large size and calculation of the YOLOv4 model, the backbone network Cross Stage Partial Darknet53 (CSPDarknet53) of the YOLOv4 model is replaced by EfficientNet, and convolution layer (Conv2D) is added to the three outputs to further adjust and extract the features, which make the model lighter and reduce the computational complexity. Finally, the apple detection experiment is performed on 2670 expanded samples. The test results show that the EfficientNet-B0-YOLOv4 model proposed in this paper has better detection performance than YOLOv3, YOLOv4, and Faster R-CNN with ResNet, which are state-of-the-art apple detection model. The average values of Recall, Precision, and F1 reach 97.43%, 95.52%, and 96.54% respectively, the average detection time per frame of the model is 0.338 s, which proves that the proposed method can be well applied in the vision system of picking robots in the apple industry.
Citations
More filters
TL;DR: In this paper, a robust real-time pear fruit counter for mobile applications using only RGB data, the variants of the state-of-the-art object detection model YOLOv4, and the multiple object-tracking algorithm Deep SORT was presented.
Abstract: This study aimed to produce a robust real-time pear fruit counter for mobile applications using only RGB data, the variants of the state-of-the-art object detection model YOLOv4, and the multiple object-tracking algorithm Deep SORT. This study also provided a systematic and pragmatic methodology for choosing the most suitable model for a desired application in agricultural sciences. In terms of accuracy, YOLOv4-CSP was observed as the optimal model, with an AP@0.50 of 98%. In terms of speed and computational cost, YOLOv4-tiny was found to be the ideal model, with a speed of more than 50 FPS and FLOPS of 6.8–14.5. If considering the balance in terms of accuracy, speed and computational cost, YOLOv4 was found to be most suitable and had the highest accuracy metrics while satisfying a real time speed of greater than or equal to 24 FPS. Between the two methods of counting with Deep SORT, the unique ID method was found to be more reliable, with an F1count of 87.85%. This was because YOLOv4 had a very low false negative in detecting pear fruits. The ROI line is more reliable because of its more restrictive nature, but due to flickering in detection it was not able to count some pears despite their being detected.
59 citations
TL;DR: In this article , a long-close distance coordination control strategy for a litchi picking robot was proposed based on an Intel Realsense D435i camera combined with a point cloud map collected by the camera.
Abstract: For the automated robotic picking of bunch-type fruit, the strategy is to roughly determine the location of the bunches, plan the picking route from a remote location, and then locate the picking point precisely at a more appropriate, closer location. The latter can reduce the amount of information to be processed and obtain more precise and detailed features, thus improving the accuracy of the vision system. In this study, a long-close distance coordination control strategy for a litchi picking robot was proposed based on an Intel Realsense D435i camera combined with a point cloud map collected by the camera. The YOLOv5 object detection network and DBSCAN point cloud clustering method were used to determine the location of bunch fruits at a long distance to then deduce the sequence of picking. After reaching the close-distance position, the Mask RCNN instance segmentation method was used to segment the more distinctive bifurcate stems in the field of view. By processing segmentation masks, a dual reference model of “Point + Line” was proposed, which guided picking by the robotic arm. Compared with existing studies, this strategy took into account the advantages and disadvantages of depth cameras. By experimenting with the complete process, the density-clustering approach in long distance was able to classify different bunches at a closer distance, while a success rate of 88.46% was achieved during fruit-bearing branch locating. This was an exploratory work that provided a theoretical and technical reference for future research on fruit-picking robots.
32 citations
TL;DR: This method takes YOLOX-Tiny as the baseline and uses the lightweight network Shufflenetv2 added with the convolutional block attention module (CBAM) as the backbone to improve the detection accuracy, and only two extraction layers are used to simplify the network structure.
Abstract: In order to enable the picking robot to detect and locate apples quickly and accurately in the orchard natural environment, we propose an apple object detection method based on Shufflenetv2-YOLOX. This method takes YOLOX-Tiny as the baseline and uses the lightweight network Shufflenetv2 added with the convolutional block attention module (CBAM) as the backbone. An adaptive spatial feature fusion (ASFF) module is added to the PANet network to improve the detection accuracy, and only two extraction layers are used to simplify the network structure. The average precision (AP), precision, recall, and F1 of the trained network under the verification set are 96.76%, 95.62%, 93.75%, and 0.95, respectively, and the detection speed reaches 65 frames per second (FPS). The test results show that the AP value of Shufflenetv2-YOLOX is increased by 6.24% compared with YOLOX-Tiny, and the detection speed is increased by 18%. At the same time, it has a better detection effect and speed than the advanced lightweight networks YOLOv5-s, Efficientdet-d0, YOLOv4-Tiny, and Mobilenet-YOLOv4-Lite. Meanwhile, the half-precision floating-point (FP16) accuracy model on the embedded device Jetson Nano with TensorRT acceleration can reach 26.3 FPS. This method can provide an effective solution for the vision system of the apple picking robot.
20 citations
20 citations
TL;DR: In this article, a DA-YOLOv7 model was proposed to detect Camellia oleifera fruit in complex field scenes, which achieved the best detection performance and a strong generalisation ability in complex scenes.
Abstract: Rapid and accurate detection of Camellia oleifera fruit is beneficial to improve the picking efficiency. However, detection faces new challenges because of the complex field environment. A Camellia oleifera fruit detection method based on YOLOv7 network and multiple data augmentation was proposed to detect Camellia oleifera fruit in complex field scenes. Firstly, the images of Camellia oleifera fruit were collected in the field to establish training and test sets. Detection performance was then compared among YOLOv7, YOLOv5s, YOLOv3-spp and Faster R-CNN networks. The YOLOv7 network with the best performance was selected. A DA-YOLOv7 model was established via the YOLOv7 network combined with various data augmentation methods. The DA-YOLOv7 model had the best detection performance and a strong generalisation ability in complex scenes, with mAP, Precision, Recall, F1 score and average detection time of 96.03%, 94.76%, 95.54%, 95.15% and 0.025 s per image, respectively. Therefore, YOLOv7 combined with data augmentation can be used to detect Camellia oleifera fruit in complex scenes. This study provides a theoretical reference for the detection and harvesting of crops under complex conditions.
20 citations
References
More filters
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
123,388 citations
Proceedings Article•
03 Dec 2012TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
73,978 citations
Proceedings Article•
04 Sep 2014TL;DR: This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.
55,235 citations
Proceedings Article•
01 Jan 2015TL;DR: In this paper, the authors investigated the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting and showed that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 layers.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.
49,914 citations
01 Jan 1998
TL;DR: In this article, a graph transformer network (GTN) is proposed for handwritten character recognition, which can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters.
Abstract: Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.
42,067 citations