Font Size: a A A

Study On The Real-time Capturing, Monitoring And Death Time Diagnosis For Insects

Posted on:2015-02-19Degree:MasterType:Thesis
Country:ChinaCandidate:J L YuFull Text:PDF
GTID:2253330425487343Subject:Agricultural Biological Environmental and Energy Engineering
Abstract/Summary:
Crop pests are not only the main factors of restricting high-yielding crops, but also the culprits of harming crops’growth and quality. Accordingly, monitoring and forecasting crop pests have become the primary issues that faced by agriculture under new situation. This paper studied the in-field insects like pieris rapae, mainly focusing on the three problems of real-time capturing, rapid detection and death time diagnosis.This study analyzed the current mainstream insect traps from both inland and abroad, summarizing their external structure, internal details of components, the application environment, methods of operation, as well as regular maintenance deeply. Moreover, this study outlined the advantages and disadvantages of the existing insect traps, and researched on a new type of insect count monitoring devices to compensate those disadvantages.The organic combination of the above two parts will contribute to the imbedding of hyperspectral imaging technology into the insect traps in future, to design the new insect-catching equipment with portable spectral identification function. The new design can accomplish the whole set of technical processes from lure, capture, counting to live-or-death identification, and death time diagnosis for insects, in order to achieve the real-time, nondestructive detection and intelligent monitoring for insects. The main contents and conclusions of this study are as follows:(1) The identification of live and dead pieris rapae using hyperspectral imaging technology. Basing on the reflection spectrum of living pieris rapae and dead pieris rapae, this study established an analysis model to judge the living state of pieris rapae. The results show that pretreatment methods have influence on the judging results, and the state of living and dead can be judged accurately. The models of PLS-DA、KNN、BPNN and SVM, which established basing on Bw and SPA choosing characteristic wavelength, all achieved good results, the accuracy of judgment is, or close to100%.(2) The diagnosis of the death time of pieris rapae using hyperspectral imaging technology. The results found that the judgment performance of Hyperspectral imaging technology is best when death time of dead pieris are within120min. While, with the increase of death time, the judgment performance are becoming worse and worse. The results indicated that it is feasible that using Hyperspectral imaging technology to identify the death time of dead pieris rapae, but it still need further study.(3) The comprehensive comparison and analysis of the domestic and overseas insect traps. In the field of pest management, in order to take proactive measures to reduce the incidence of pests, the time of insect outbreaks are needed to forecast accuracy. Moreover, in the aftermath of the pest, insect traps of high-performance are utilized to kill pests, minimizing losses caused by pests. The results showed that nowadays, both domestic and abroad mainstream insect traps works well, effective trapping rate is extremely high. While there are still some shortcomings when monitoring and capturing insects in time.(4) The design and development of a new monitoring device for real-time insect counting. In order to recover the disadvantages of existing insect traps on real-time monitoring and insect counting, this studying used matrix key influence groups innovatively, and accurately identified the behavior of insect flapping lamp instantaneously, and furthermore completed the real-time counting of living insects. The results indicated that, under theoretical derivation, the statistics of insect flapping lamp counted by the device can reach to design values, which demonstrated indirectly the ability of the insect monitoring device on counting.
Keywords/Search Tags:Hyperspectral imaging technology, Pieris rapae, Capture insectdevice and equipment, Successive projections algorithm, Insect count monitoringdevices
Related items