{"title":"电荷耦合器件(CCD)光电信号数据采集及其在人工智能机器视觉中的应用","authors":"Yan Liu, Jianhang Zeng","doi":"10.1166/jno.2023.3450","DOIUrl":null,"url":null,"abstract":"The development of Charge Couple Device (CCD) technology is particularly rapid in the fields of image sensors and non-contact measurement. In this study, a data acquisition device applied to CCD photoelectric detection system is designed. Among them, the design of the Differential Amplification (DA) module, Analog-to-Digital Converter (ADC) module, First In First Out (FIFO) cache module, and Complex Programmable Logic Device (CPLD) module in this device are emphasized. The ADC circuit in the ADC module converts two 4 MHz analog photoelectric signals generated by the CCD sensor at a frequency of 8 MHz, and then outputs 12-bit digital signals. The collected photoelectric signal is used to detect the damage to the surface of ancient buildings with the machine vision technology of artificial intelligence (AI). In the test, the DA circuit can adjust the voltage range of two photoelectric analog signals output by CCD to a predetermined range (1.5 V∼2.0 V). In the ADC circuit test, there is no data in the FIFO when there is no input conversion, and the converted data will be stored in the internal FIFO during the conversion clock period. Based on machine vision technology, surface damage types of ancient buildings are defined, namely spalling, cracks, and disruption, and surface image samples are generated from collected signals. The samples are trained using the convolutional neural network, and the classifier is generated. The test reveals that the designed photoelectric signal acquisition device and AI machine vision technology can accurately classify the surface damage of ancient buildings.","PeriodicalId":16446,"journal":{"name":"Journal of Nanoelectronics and Optoelectronics","volume":"83 1","pages":"0"},"PeriodicalIF":0.6000,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Charge Couple Device (CCD) Photoelectric Signal Data Acquisition and Its Application in the Machine Vision of Artificial Intelligence\",\"authors\":\"Yan Liu, Jianhang Zeng\",\"doi\":\"10.1166/jno.2023.3450\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The development of Charge Couple Device (CCD) technology is particularly rapid in the fields of image sensors and non-contact measurement. In this study, a data acquisition device applied to CCD photoelectric detection system is designed. Among them, the design of the Differential Amplification (DA) module, Analog-to-Digital Converter (ADC) module, First In First Out (FIFO) cache module, and Complex Programmable Logic Device (CPLD) module in this device are emphasized. The ADC circuit in the ADC module converts two 4 MHz analog photoelectric signals generated by the CCD sensor at a frequency of 8 MHz, and then outputs 12-bit digital signals. The collected photoelectric signal is used to detect the damage to the surface of ancient buildings with the machine vision technology of artificial intelligence (AI). In the test, the DA circuit can adjust the voltage range of two photoelectric analog signals output by CCD to a predetermined range (1.5 V∼2.0 V). In the ADC circuit test, there is no data in the FIFO when there is no input conversion, and the converted data will be stored in the internal FIFO during the conversion clock period. Based on machine vision technology, surface damage types of ancient buildings are defined, namely spalling, cracks, and disruption, and surface image samples are generated from collected signals. The samples are trained using the convolutional neural network, and the classifier is generated. The test reveals that the designed photoelectric signal acquisition device and AI machine vision technology can accurately classify the surface damage of ancient buildings.\",\"PeriodicalId\":16446,\"journal\":{\"name\":\"Journal of Nanoelectronics and Optoelectronics\",\"volume\":\"83 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.6000,\"publicationDate\":\"2023-07-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Nanoelectronics and Optoelectronics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1166/jno.2023.3450\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Nanoelectronics and Optoelectronics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1166/jno.2023.3450","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
摘要
电荷耦合器件(CCD)技术在图像传感器和非接触式测量领域的发展尤为迅速。本课题设计了一种应用于CCD光电检测系统的数据采集装置。其中,着重介绍了该器件中差分放大(DA)模块、模数转换器(ADC)模块、先进先出(FIFO)缓存模块和复杂可编程逻辑器件(CPLD)模块的设计。ADC模块中的ADC电路将CCD传感器产生的2个频率为8mhz的4mhz模拟光电信号进行转换,然后输出12位数字信号。利用人工智能的机器视觉技术,将采集到的光电信号用于古建筑表面的损伤检测。在测试中,DA电路可以将CCD输出的两个光电模拟信号的电压范围调整到预定范围(1.5 V ~ 2.0 V),在ADC电路测试中,当没有输入转换时,FIFO中没有数据,转换后的数据将在转换时钟周期内存储在内部FIFO中。基于机器视觉技术,定义古建筑的表面损伤类型,即剥落、裂缝和破坏,并根据采集到的信号生成表面图像样本。使用卷积神经网络对样本进行训练,生成分类器。试验表明,所设计的光电信号采集装置和人工智能机器视觉技术能够对古建筑表面损伤进行准确分类。
Charge Couple Device (CCD) Photoelectric Signal Data Acquisition and Its Application in the Machine Vision of Artificial Intelligence
The development of Charge Couple Device (CCD) technology is particularly rapid in the fields of image sensors and non-contact measurement. In this study, a data acquisition device applied to CCD photoelectric detection system is designed. Among them, the design of the Differential Amplification (DA) module, Analog-to-Digital Converter (ADC) module, First In First Out (FIFO) cache module, and Complex Programmable Logic Device (CPLD) module in this device are emphasized. The ADC circuit in the ADC module converts two 4 MHz analog photoelectric signals generated by the CCD sensor at a frequency of 8 MHz, and then outputs 12-bit digital signals. The collected photoelectric signal is used to detect the damage to the surface of ancient buildings with the machine vision technology of artificial intelligence (AI). In the test, the DA circuit can adjust the voltage range of two photoelectric analog signals output by CCD to a predetermined range (1.5 V∼2.0 V). In the ADC circuit test, there is no data in the FIFO when there is no input conversion, and the converted data will be stored in the internal FIFO during the conversion clock period. Based on machine vision technology, surface damage types of ancient buildings are defined, namely spalling, cracks, and disruption, and surface image samples are generated from collected signals. The samples are trained using the convolutional neural network, and the classifier is generated. The test reveals that the designed photoelectric signal acquisition device and AI machine vision technology can accurately classify the surface damage of ancient buildings.