Activation functions are crucial in deep learning networks, given that the nonlinear ability of activation functions endows deep neural networks with real artificial intelligence. Nonlinear nonmonotonic activation functions, such as rectified linear units, Tan hyperbolic (tanh), Sigmoid, Swish, Mish, and Logish, perform well in deep learning models; however, only a few of them are widely used in mostly all applications due to their existing inconsistencies. Inspired by the MB-C-BSIF method, this study proposes Smish, a novel nonlinear activation function, expressed as $f\left(x\right)=x·tanh\left[\ln \left(1+sigmoid\left(x\right)\right)\right]$, which could overcome other activation functions with good properties. Logarithmic operations are first used to reduce the range of sigmoid(x). The value is then calculated using the tanh operator. Inputs are ultimately used to multiply the previous value, thus exhibiting negative output regularization. Experiments show that Smish tends to operate more efficiently than Logish, Mish, and other activation functions on EfficientNet models with open datasets. Moreover, we evaluated the performance of Smish in various deep learning models and the parameters of its function $f\left(x\right)=\alpha x·tanh\left[\ln \left(1+sigmoid\left(\beta x\right)\right)\right]$, and where α = 1 and β = 1, Smish was found to exhibit the highest accuracy. The experimental results show that with Smish, the EfficientNetB3 network exhibits a Top-1 accuracy of 84.1% on the CIFAR-10 dataset; the EfficientNetB5 network has a Top-1 accuracy of 99.89% on the MNIST dataset; and the EfficientnetB7 network has a Top-1 accuracy of 91.14% on the SVHN dataset. These values are superior to those obtained using other state-of-the-art activation functions, which shows that Smish is more suitable for complex deep learning models.

We propose a new image encryption algorithm based on DNA sequences combined with chaotic maps. This algorithm has two innovations: (1) it diffuses the pixels by transforming the nucleotides into corresponding base pairs a random number of times and (2) it confuses the pixels by a chaotic index based on a chaotic map. For any size of the original grayscale image, the rows and columns are fist exchanged by the arrays generated by a logistic chaotic map. Secondly, each pixel that has been confused is encoded into four nucleotides according to the DNA coding. Thirdly, each nucleotide is transformed into the corresponding base pair a random number of time(s) by a series of iterative computations based on Chebyshev’s chaotic map. Experimental results indicate that the key account of this algorithm is 1.536 × 10127, the correlation coefficient of a 256 × 256 Lena image between, before, and after the encryption processes was 0.0028, and the information entropy of the encrypted image was 7.9854. These simulation results and security analysis show that the proposed algorithm not only has good encryption effect, but also has the ability to repel exhaustive, statistical, differential, and noise attacks.

Forest fires represent a real threat to human lives, ecological systems, and infrastructure. Many commercial fire detection sensor systems exist, but all of them are difficult to apply at large open spaces like forests because of their response delay, necessary maintenance needed, high cost, and other problems. In this paper a forest fire detection algorithm is proposed, and it consists of the following stages. Firstly, background subtraction is applied to movement containing region detection. Secondly, converting the segmented moving regions from RGB to YCbCr color space and applying five fire detection rules for separating candidate fire pixels were undertaken. Finally, temporal variation is then employed to differentiate between fire and fire-color objects. The proposed method is tested using data set consisting of 6 videos collected from Internet. The final results show that the proposed method achieves up to 96.63% of true detection rates. These results indicate that the proposed method is accurate and can be used in automatic forest fire-alarm systems.

With the development of national information processes, specific image information from secret departments or individuals is often required to be confidentially transmitted. Numerous image encryption methods exist, especially since the initial value sensitivity and other characteristics of chaos theory and chaos theory-based encryption have become increasingly important in recent years. At present, DNA coding constitutes a new research direction of image encryption that uses the four base pairs of DNA code and image pixel values to establish a special correspondence, in order to achieve pixel diffusion. There are eight DNA encoding rules, and current methods of selecting the DNA encoding rules are largely fixed. Thus, the security of encoded data is not high. In this paper, we use the Lorenz chaotic system, Chen’s hyperchaotic system, and the DNA encoding combination and present a new image encryption algorithm that can dynamically select eight types of DNA encoding rules and eight types of DNA addition and subtraction rules, with significant improvements in security. Through simulation experiments and histograms, correlations, and NPCR analyses, we have determined that the algorithm possesses numerous desirable features, including good encryption effects and antishear and antinoise performances.

Small object detection is one of the challenging tasks in computer vision. Most of the existing small object detection models cannot fully extract the characteristics of small objects within an image, due to the small coverage area, low resolution and unclear detailed information of small objects in the image; hence, the effect of these models is not ideal. To solve this problem, a simple and efficient reinforce feature pyramid network R-FPN is proposed for the YOLOv5 algorithm. The learnable weight is introduced to show the importance of different input features, make full use of the useful information of different feature layers and strengthen the extraction of small object features. At the same time, a channel space mixed attention CSMA module is proposed to extract the detailed information of small objects combined with spaces and channels, suppress other useless information and further improve the accuracy of small object detection. The experimental results show that the proposed method improves the average accuracy AP, AP50 and AR100 of the original algorithm by 2.11%, 2.86% and 1.94%, respectively, and the detection effect is better than the existing small object detection algorithms, which proves the effectiveness of the proposed method.

With the wide increase in global forestry resources trade, the demand for wood is increasing day by day, especially rare wood. Finding a computer-based method that can identify wood species has strong practical value and very important significance for regulating the wood trade market and protecting the interests of all parties, which is one of the important problems to be solved by the wood industry. This article firstly studies the establishment of wood microscopic images dataset through a combination of traditional image amplification technology and Mix-up technology expansion strategy. Then with the traditional Faster Region-based Convolutional Neural Networks (Faster RCNN) model, the receptive field enhancement Spatial Pyramid Pooling (SPP) module and the multi-scale feature fusion of Feature Pyramid Networks (FPN) module are introduced to construct a microscopic image identification model based on the migration learning fusion model and analyzes the three factors (Mix-up, Enhanced SPP and FPN modules) affecting the wood microscopic image detection model. The experimental results show that the proposed approach can identify 10 kinds of wood microscopic images, and the accuracy rate has increased from 77.8% to 83.8%, which provides convenient conditions for further in-depth study of the microscopic characteristics of wood cells and is of great significance to the field of wood science.