The increasing perfection of BeiDou positioning system has increased the market of application terminals, but positioning accuracy has become a major problem for users. According to the advantages of Kalman Filtering algorithm, such as time continuity and noise reduction, this paper briefly introduces the principle of pseudo range positioning, introduces, simulates and analyzes the error of least square method and Kalman Filtering method, and proposes a traceless Kalman Filtering algorithm to improve the least square method for adaptive noise estimation. Through MATLAB simulation, the advantages and disadvantages of the algorithm performance are compared, and the feasibility of the algorithm in improving the positioning accuracy is verified.
At present, our country's plastic industry is developing rapidly. When plastic products bring great convenience to people's lives, waste plastic waste also causes huge pollution to the environment. Sorting, in view of the problem that various colors of plastic bottles are mixed in the sorting of waste plastic bottles, and the sorting is more labor-intensive, which leads to the increase of recycling costs. This paper studies the denoising process and color feature extraction before image recognition of plastic bottles, using HSV The model replaces the RGB model for color feature extraction. The recognition rate of the HSV model is 96.67%, and the recognition rate of the RGB model is 95.00%, but the recognition speed of HSV is increased by 16.43%.
In order to solve the problems of low detection accuracy and poor robustness of fruits under natural conditions, an apple detection model based on improved yolov3 is proposed with apples in natural environment as the research object. Firstly, the ordinary convolution inside the residual network in DarkNet53 is replaced with a depth-separable convolution to improve the detection speed of the model while reducing the number of parameters in the network; secondly, the channel-space attention mechanism (Convolutional Block Attention Module, CBAM)module is added to the process of feature pyramid upsampling to further strengthen the model feature extraction and improve the recognition accuracy of overlapping apples; finally, the introduction of CIOU Loss complete cross-comparison loss function, which integrates the influence of the overlapping area, aspect ratio, centroid distance and other factors on the bounding box regression, makes the bounding box regression stable while the target detection accuracy is also improved. The experiments show that the improved model achieves a MAP (Mean Average Precision) value of 91.30% and an F1 value of 86% under the test set, which is 4.2 percentage points and 5 percentage points hight than the original model respectively.
With the rapid development of automation technology, the processing of various kinds of shaped parts has received the attention of the majority of scholars, such as the identification of automotive door panel weld joints. Automotive door plate weld joint features are small targets, interference, and uneven distribution. The previous solder joint recognition method is affected by the small target and many interferences, and these disturbances lead to inaccurate solder joint identification. To address the problem of inaccurate solder joint recognition, a solder joint recognition method based on the improved YOLOv5 algorithm is proposed. Firstly, the dataset of solder joints is constructed, and then the YOLOv5 network structure is improved by adding a tiny target detection layer,which makes the accuracy of small target detection improved. The accuracy of the improved YOLOv5 for solder joint detection is improved from the original 91.7% to 98.9%, which can be extended to such small target application scenarios.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.