Aiming at the problems that a wide variety of Chinese medicinal materials have small samples, and it is difficult to classify the vessels of them, an improved convolutional neural network method was proposed based on multi-channel color space and attention mechanism model. Firstly, the multi-channel color space was used to merge the RGB color space with other color spaces into 6 channels as the network input, so that the network was able to learn the characteristic information such as brightness, hue and saturation to make up for the insufficient samples. Secondly, the attention mechanism model was added to the network, in which the two pooling layers were connected tightly by the channel attention model, and the multi-scale cavity convolutions were combined by the spatial attention model, so that the network focused on the key feature information in the small samples. Aiming at 8 774 vessel images of 34 samples collected from Chinese medicinal materials, the experimental results show that by using the multi-channel color space and attention mechanism model method, compared with the original ResNet network, the accuracy is increased by 1.8 percentage points and 3.1 percentage points respectively, and the combination of the two methods increases accuracy by 4.1 percentage points. It can be seen that the proposed method greatly improves the accuracy of small-sample classification.
To solve the problem of discontinuity when blending two surfaces with coplanar perpendicular axis, this paper discussed how to improve the equations about the blending surface so as to obtain the smooth and continuous blending surface. At first, this paper analyzed the reason of the uncontinuousness in the blending surface and pointed out that the items in one variable were removed when other variables equaled to some specified values. In this case, the blending equation was independent to this variable in these values and this indicated that the belending surface was disconnected. Then, a method which guarantees the blending surface countinuous was presented on the basis of above discussion. Besides this, this paper discussed how to smoothen it once the continuous blending surface was computed out. As for the G0 blending surface, regarding the polynomial of auxiliary surface as a factor, this factor was mulitiplied to a function f′ with degree one and the result was added to the primary surface fi. The smoothness of blending surface can be implemented by changing the coefficients in f. For the Gn blending surface, a compensated polynomial with degree at most 2 was added to the proposed primary blending equation directly when computing blending surface. This method smoothens the blending surface but does not increase the degree of G0 blending surface.
An improved method for Multi-Scale Retinex with Color Restoration (MSRCR) algorithm was proposed, to remove the fog at the far prospect and solve gray hypothesis problem. First, original fog image was inverted. Then, MSRCR algorithm was used on it. The inverted image was to be inverted again and then was linearly superposed with the result which was processed by MSRCR algorithm directly .At the same time , the reflection component which was got during the process of the extraction was linearly superposed with the original luminance, and the mean and variance were calculated to decide the contrast stretching degree adaptively, finally, it was uniformly stretched to the display device.The experimental results show that the proposed algorithm can get a better effect of removing the fog. Evaluation values of the processed image, including standard difference, average brightness, information entropy, and squared gradient, are improved than the original algorithm. It is easy to implement and has important significance for real-time video to remove fog.