Search

The symptoms of the same tomato disease in different stages are obviously different, and different diseases show some similarities. Traditional pattern recognition methods can not reflect the dynamic changes of the pathological characterization, and the practicability is poor. To solve this problem, this paper proposed a Multi-Scale AlexNet recognition model for mobile platform based on convolutional neural network (CNN), and implemented a tomato leaf disease image recognition system for agricultural workers based on Android. Many parameters and large memory utilization of traditional AlexNet model are unfit for mobile platform, this paper adjusted the network structure of the traditional model by removing the local response normalization(LRN) layer, modifying the full connection layer, setting up different convolution kernel extraction features, designed a multi- scale recognition model based on the AlexNet. The model consists of 6 layers. It can optimize the training time and memory utilization and achieve high precision. After removing the LRN layer, there was a 30% decrease in running time. Extending the single convolution kernel into multi-scale (1 × 1,3 × 3,5 × 5,7 × 7) convolution kernel then fused at the first layer, removing full connection layer 6 and 7, and taking the place of global average pooling layer, then the model size was only 30.2 M. The forward propagation rate (FRP) and backward propagation rate (BPR) were reduced, and the global average pooling is better than the global maximum pooling on recognition accuracy. So the Multi-Scale AlexNet model used global average pooling in the 5th layer. In image preprocessing phase, in order to avoiding over fitting of the trained model caused by the unbalanced distribution of sample numbers, we had zoomed, flipped, color jittering, add noise and rotated the original pictures of dataset randomly to get the augmented dataset, and used 70% of the pictures as the train dataset and the rest as the validation dataset(20%) and test dataset(10%). These pictures were quantized to 224 × 224 dpi for Multi-Scale AlexNet training, and the original dataset and augmented dataset were used to train models. In order to validate the performance of the proposed model, comparative tests were done between Multi-Scale AlexNet and traditional pattern recognition method. It repeated 600 tests. The results showed that the CNN model achieved 92.7%, the high average recognition accuracy of each disease and each disease in the early, middle and late stages. Compared with the other CNN Net model(MobileNet, SequeezeNet, LeNet-5), the Multi-Scale AlexNet achieved the highest recognition accuracy, and reached 95.8% on the late stage disease dataset, but the SequeezeNet model used less memory. The MobileNet and SequeezeNet model reached lower recognition accuracy on the middle and late stages dataset, that because their convolution size was small. The recognition system was implemented on Android platform, and then test was done on field dataset. The results showed that the average recognition accuracy was 89.2%, its less value was due to the complex background of image. Then the system can meet the requirements of disease image recognition on mobile platform in production practice. The research results provide a method for disease image recognition based on convolution neural network, and provide a reference for automatic identification of crop diseases and engineering applications. [ABSTRACT FROM AUTHOR]

Based on the whole process of malware detection in the Android system. this paper compared and analyzed the status and progress of the research at home and abroad, introduced the source and function of the standardized data samples from the point of view of sample acquisition, and expounded the principles to be followed in the feature selection. This paper compared and analyzed the advantages and disadvantages of various detection methods from the point of view of detection methods, and summed up and summarized the selection methods of feature data sets and the evaluation methods of experimental results. Finally, it prospected the research and development direction of the future Android malware detection method. [ABSTRACT FROM AUTHOR]

Due to the openness of Android system, malware poses a threat to users of Android devices by implementing various malicious behaviors. At present, most of the existing researches focus on coarse-grained malicious detection, that is, whether an Android application is malicious or not. Aiming at this problem, this paper proposed a fine-grained malicious behavior classification method based on static behavior features. This method extracted and optimized multi-dimensional behavior characteristics, including API calls, permissions, intents and package dependencies. And then used random forest to classify malicious behaviors. It conducted the experiments on 24 553 malicious Android application samples in 73 malware families from multiple application markets. The experimental results show that the accuracy of fine-grained malware classification reached to 95. 88%, which is better than other comparison methods. [ABSTRACT FROM AUTHOR]

Many Android applications have more functions than their types, and they need to acquire more permissions. Excessive permissions may bring some security risks. To address these issues, this paper proposed an Android malware detection method based on meta information. First, it extracted the LDA theme through the description of Android application, implemented the data dimensionality reduction, and grouped applications by the functional type used the K-means clustering al gorithm. Then, for all applications belonging to the same functional type, it extracted their permission information, and took the permission features as the research object, used KNN algorithm to classify and detect the malicious software of Android. The experimental results obtain the average accuracy of 94. 81 % and prove the validity and high accuracy of the method. [ABSTRACT FROM AUTHOR]

Due to the shortage of traditional door assessment system, such as single function, poor reliability and low security, a multi-door access control system based on fingerprint identification technology is put forward in this paper. HT66FU50 is selected as the control core of this system, it mainly adopts fingerprint recognition technology to control two or more doors. When the system gets illegal fingerprints for three times, it will send a message and ring a call to the telephone of the owner through the GSM. Meanwhile it connects the camera by Wi-Fi so that it can capture image and video. PC and Android phone can also monitor and control the situation through Wi-Fi technology, so that the access control system is safer and more reliable. [ABSTRACT FROM AUTHOR]