Search

This paper researches on bacon quality pattern recognition. We take the hyper spectral microscopic image of plaque area, the acid value, the peroxide value (POV) and total volatile basic nitrogen (TVBN) obtained from the spectra at the Self Organization Map (SOM) neural network as inputs to conduct multi-data fusion. From that we set up a model about level detection of bacon quality. Eventually we refine the level of bacon quality to four categories based on the national standard with rough division of 2 classes, Those are: Safe to eat, Edible, Do not recommend eating and Inedible, which increase the detection accuracy of pattern recognition, and can propose an early warning for eating bacon safely. The study found that the detection model's root mean square prediction error (RMSEP) is respectively 3%, 9%, 4% of acid value, peroxide value and TVBN built after pretreatment and band selection. The number of wavelengths used here is about one-third of the whole number of wavelengths, which has improved the detection accuracy of pattern recognition. Theories and methods of the research can be further extended to detect other related agricultural meat products.

In this paper a quick and accurate detection method is proposed, which can identify whether the sausage contains excessive acid value, peroxide value and area of colony. By using hyperspectral image measurement and multi-information fusion based on support vector machine (SVM), the sausage content model is established. In order to improve the accuracy of hyperspectral image measurement predicted model and to reduce the measurement turbulence, the image information and the NIR value data that are input as the parameters of the hyperspectral image content model are introduced. The detection model's RMSECV and r2 of the research are 0.251 and 0.972. The study concludes that the theory and method can be further extended to the detection of other related meat agricultural products.

For the control of automobile airbag ignition, this paper focused on algorithms based on BP neural network. First, some common airbag ignition algorithms and their deficiencies were introduced, and then algorithms based on BP neural network were discussed. The main design and simulation were introduced, and the results show that the algorithm based on BP neural network is a good solution.

Infrared and visible image fusion technology can effectively improve the image contrast and clarity, and enhance the night vision effective. Non-sub sampled contourlet transform (NSCT) in image fusion field has made some achievements. A regional standard deviation-weighted image fusion method based on non-sub sampled Contourlet transform was proposed, and the robustness of the method was analyzed. First, the reiterated infrared and visible images from the same scene were transformed by non-sub sampled contourlet transform, followed the approximate weight averaged, high-frequency detail components in accordance with the weighted of the regional standard deviation proportion, then the fusion image is obtained by inverse non-sub sampled contourlet transform, Finally, the fusion images were compared with the results obtained by Laplace transform, wavelet transform and contourlet transform through a large number of experiments, and the robustness analysis was done through the noise test. The results show that: non-sub sampled contourlet transform can achieve better fusion effect, and high robustness.