Search

With the advancement of remote sensing technology, the high spatial resolution remote sensing image contains rich special information with a great detail. At the same time, the complexity of high spatial resolution remote sensing images also requires higher the classification technology of remote sensing images. However, in the face of high spatial resolution remote sensing image more obvious geometrical structure and the more rich texture characteristics, how to design rational system of characteristics, select the appropriate sorting algorithms to accurately and quickly grasp the number of rural land of building and its distribution status, are of great significance to balance urban and rural areas, save land, and realize sustainable development. This will help in exploring the application of deep learning model in high spatial resolution remote sensing image building extraction, and have research significance for improving the classification accuracy of high resolution remote sensing image. In this paper, the semantic segmentation model (SegNet) was used for extracting buildings. SegNet is mainly composed of encoder network, decoder network and pixel-wise classification layer. The encoder network transforms high-dimensional vectors into low-dimensional vectors, enabling low-dimensional extraction of high-dimensional features. The decoder network maps low-resolution feature maps to high spatial resolution feature maps, realizing the reconstruction of low-dimensional vectors to high-dimensional vectors. The softmax classifier separately classifies each pixel, which outputs the probability that each pixel belongs to each class. In this paper, a 3000 pixel × 3000 pixel and two 2000 pixel × 2000 pixel slices were taken from the global remote sensing image of Bazhou City, Hebei Province as training samples, and a 3000 pixel × 3000 pixel slice was taken as the verification sample. In this paper, five comparative experiments were used to extract the buildings, including PSPNet, support vector machine, random forest, ISO clustering and maximum likelihood method. The confusion matrix of each classification method was obtained by calculating the difference between the classification results of the comparison experiment and the real value. From the traditional classification algorithm to the shallow learning algorithm to the deep learning algorithm, the Kappa coefficient and overall accuracy of classification kept constantly increasing, among which SegNet semantic segmentation algorithm based on the deep convolutional network performed better than the other five algorithms in extracting buildings from high spatial resolution remote sensing image. The Kappa coefficient and the overall accuracy of SegNet semantic segmentation algorithm were 0.90 and 96.61%, respectively, and the ground truth value was basically the same as the classification result. The F1Score of building extraction of SegNet semantic segmentation algorithm based on deep convolution network was 0.91, but the other five algorithms were below 0.87. SegNet had the lowest error rate of 9.71% for buildings, indicating that the ability to identify buildings of semantic segmentation algorithm from high spatial resolution remote sensing was superior to traditional classification algorithms, shallow layer learning algorithms based on machine learning, and PSPNet semantic segmentation algorithm based on deep convolution network. The Kappa coefficient and overall accuracy of the remaining five classification algorithms were respectively below 0.83 and 94.68%, and the difference between the ground truth value and the classification result was relatively large. SegNet can not only make use of spectral information but also make full use of abundant spatial information. During SegNet training, more essential features can be learned, and more ideal features suitable for pattern classification were finally formed, which can enhance the ability of convergence and generalization of the model and improve the classification accuracy. Traditional classification algorithms, such as ISO clustering and maximum likelihood method, failed to make use of the rich spatial information of the high-resolution remote sensing image, so the accuracy was relatively low. Due to limited computing units and large amount of high spatial resolution remote sensing image data, shallow layer learning algorithms based on machine learning such as support vector machines and random forest cannot effectively express complex features of ground objects, so their advantages are not obvious in building extraction from the high spatial resolution remote sensing images. The experimental results showed that the SegNet based on deep learning has the best performance, and it has important theoretical significance to explore the application of deep learning model to remote sensing image classification methods. At the same time, the research results also provide a reference for improving the classification accuracy of high resolution remote sensing images. [ABSTRACT FROM AUTHOR]

Timely and accurate identification of oestrus behavior of sows can effectively increase the conception rate and litter size, which is of great significance to improve the breeding level and economic benefits of breeding enterprises.The traditional identification methods of oestrus behavior of sows have the problems of high working intensity, strong subjectivity and low automation level. Some identification methods of oestrus behavior of sows based on internet of things also have some problems, such as high false alarm times, high error rate and long recognition time. In order to solve the problems of low identification accuracy and efficiency, a oestrus behavior recognition method of sows based on Moth-Flame Optimization optimized Long Short Term Memory(MFO-LSTM) was proposed in this paper, and the verification test was carried out at WuFeng breeding farm in Xinghualing district, Taiyuan city, Shanxi Province. The posture data of sows were obtained by posture sensors installed on the neck of the sows. The posture data describing the sow's motion state contained three-axis acceleration, three-axis angular velocity, three-axis angle, three-axis magnetic field and quaternion. The collected posture data was manually marked according to the videos, and preprocessed to obtain a posture classification data set. Then, the MFO algorithm was used to optimize the numbers of first and second hidden layer neurons, maximum training period, block size and learning rate, thus, the LSTM network model was built. Postures of sows were divided into three categories, i.e. standing, lying and mounting. Through the statistics of the duration of each complete mounting behavior of the sows, the range of the recognition time of the oestrus behavior was obtained. The sows posture classification results were counted with different oestrus behavior recognition duration, and two characteristics of mounting behavior and activity were then extracted, so as to obtain the recognition feature matrix of oestrus behavior of sows. Finally, the feature matrix was input into the MFO-LSTM classification model to judge whether the sow was oestrus. The experimental results showed that the classification method of sows posture proposed could effectively distinguish the three postures of standing, lying and mounting. The classification effects of proposed method were better than that of the Support Vector Machine (SVM), Probabilistic Neural Network(PNN), Learning Vector Quantization (LVQ) and Extreme Learning Machine(ELM). The average accuracy, recall rate and F1 in the attitude data set were 98.02%, 96.26% and 96.18, respectively. On the basis of accurately identifying the sows’ posture, the effect of oestrus behavior recognition based on MFO-LSTM algorithm was verified. The test results showed that the recognition effect was best when the oestrus recognition duration was 30 min at this time, the error rate, recall rate and specificity of oestrus recognition were 13.43%, 90.63% and 81.63%, respectively. The oestrus behavior of sows recognition method proposed effectively reduced the error rate while maintaining a high recall rate and specificity. Compared with other methods, the error rate was reduced by more than 80%, the oestrus behavior of sow could be recognized after 30 min of oestrus. [ABSTRACT FROM AUTHOR]

Traditional methods for classifying crop pests are based on the appearance of the pests such as their color, morphology and texture using algorithms s back propagation neural network and support vector machine. These methods are sensitive to the environments where the pests appear. Also, the imbalance between training sample numbers, difference in scales at which the training samples are taken, along with that the pests might seasonally change their color and shape, make it difficult for these algorithms to detect and recognize the pests. In order to improve the accuracy of detecting crop pests appearing in diverse environments, this paper proposed a deep convolutional neural network method by combining the spatial pyramid pooling with the improved YOLOv3 deep convolutional neural network algorithm. The proposed method first located the pests on a test image and then identified the species they belongs to. Using the improved network structure in YOLOv3, the proposed method used up-sample and convolution operations to calculate the de-convolution. Bilinear interpolation was used to enhance the output of the network, and the depth characteristics of the pests were extracted by the depth residual neural network. As a result, the improved network can effectively detect and recognize small pests in the images. By fusing with the spatial pyramid pooling, the Yolov3-SPP network can map the extracted eigenvectors to different spatial dimensions. The network can be used at various scales to detect pests of different sizes. Results from identifying and testing 20 types of pests collected by traps showed that the average accuracy of the proposed method was 88.07%, with a detection speed of 26 frames/s. Compared with the typical YOLOv3 algorithm, the proposed methods improved accuracy by 2.8 percentage points. We also reconstructed the maps of the network at each stage and compared them with the typical YOLOv3 algorithm in attempts to demonstrate that the features extracted by the proposed method was more recognizable. The results revealed that the typical YOLOv3 algorithm could miss the targeted objects, while the proposed method is not only able to detect pests of different sizes, it also has high recognition accuracy. Using the same dataset, we compared the proposed method with other detection algorithms such as HOG+SVM, Faster R-CNN, YOLOv3. The comparison showed that the proposed method was 19.61 percentage points higher than that of HOG + SVM, and 9.64 percentage points higher than that of Faster R-CNN, Faster R-CNN was unable to extract small pests, and its recognition accuracy was only 78.43%. [ABSTRACT FROM AUTHOR]

During the detecting and locating process of green tomatoes based on obtained visible images, problems such as shadow caused by uneven illumination source, occlusion of stems and leaves, and occlusion between fruits, need to be solved. In this study, a machine vision algorithm was put forward, which aimed to determine fruit location and size of green tomatoes. Normalized cross-correlation function (NCC) is a feature detection method using template matching. The algorithm firstly detected the potential location of green tomatoes through fast normalized cross-correlation function (FNCC). Then a gray histogram based classifier was used to classify if the location corresponded to green fruit. The histogram based classifier was built based on fruit areas and non-fruit areas of a certain size (30×30 pixels) extracted from the obtained image. Seven texture features, including mean, standard deviation, smoothness, third moment, uniformity, entropy, and gray level range were calculated for those fruit areas and non-fruit areas. Three classifiers including k-nearest neighbor (KNN), SVM (support vector machine), and Naive Bayes, were used to classify fruit and non-fruit areas using those 7 texture features as input vectors. SVM was chosen based on its performance. The non-fruit location was filtered out, and the number of fruit locations was estimated. Meanwhile, the image was segmented based on color analysis. Red and Blue component from RGB (red, green, blue) image, and Hue component from HSV (hue, saturation, value) image transformed from RGB images, were used as the basis for color analysis. Using the fruit potential location number estimated using FNCC as an input parameter of circular Hough transform (CHT), CHT was then applied to the edge image of the segmented result. The center coordinates and radius value of each circle were calculated. Finally, the detection results were merged based on the analysis of the distances of 2 centers of fruit circles detected using CHT. If the distance between 2 circles is smaller than the minimum fruit radius, the circle with a larger radius will be kept, while the other circle will be flagged as repeatedly detected one. Thus, the recognition and positioning of the green tomato were realized. When green fruits and red fruits appear on the same image, a red fruit detection algorithm based on the local maximum value method and random circle round transform detection, which was proposed by the author in another paper, would be carried out on the obtained image. Then the red fruit detection result was combined with the green fruit detection result. The proposed method combined the texture, color, and shape information of the image, and eliminated the disturbance of the color similarity between the green tomatoes and green leaves and stems. A total of 70 images were used in this study, in which 35 images were used as training images, and the other 35 images were used as validation images. The correct detection ratio for 72 fruits in the training dataset was 89.2%, and the correct detection ratio for 105 fruits in the validation dataset was 86.7%. The proposed method has provided a reference for the development of tomato harvesting robots for both red and green mature tomato fruits. [ABSTRACT FROM AUTHOR]