The paper presents an approach for medical image categorization based-on Gaussian mixture model in CBMIR system. The medical image categorization is a very complicated problem because the characteristics on texture, shape and intensity among the images of different parts of body are distinct differences. First, we extract the characteristic vectors of the training image set. Then, we choose the optimum features which can distinguish different classes and the same class better. After getting GMM parameters by EM algorithm, we categorize the test images. The experimental results indicate that the method performs well on CT image categorization.
The use of sparse representations in signal and image processing is gradually increasing in the past several years. Considering the relativity among the multi-polarimetric SAR images, a new compression scheme for multi-polarimetric SAR image based sparse representation and super-resolution is proposed. Multilevel dictionary is learned iteratively in the 9/7 wavelet domain using one channel SAR image, and the other channels are compressed by down-sampling followed with sparse representation scheme. The super-resolution algorithm is used to restore the high frequency information removed during the down-sampling process. Experimental results are compared with state-of-the-art compression methods JPEG2000, and our method performs well in terms of both subjective quality and edge preservation.
This paper presents AutoMLPMixer which is a lightweight backbone based on the mixer-multi-layer percep-trons(MLPMixer) and automated machine learning. Secondly, in order to deal with parameter surging caused by a large dictionary in Chinese scene text recognition, this paper introduces grouped linear structure to process the embedding layers and prediction layers of the model. Compared with ResNet, the AutoMLPMixer can effectively reduce parameters by 16.4%. By introducing a grouped linear structure to solve the oversized dictionary problem, we can further reduce parameters by 14.89M in scene text recognition. Under coexist condition of both AutoMLPMixer and grouped linear structure, the model image processing speed increased by 29%. The code and models will be made publicly.
Influence of different plant density on physiological properties and yield of maize was studied by field experiment under the ecological conditions of the Southern Hebei.The results showed that chlorophyll content,photosynthesis rate and soluble protein content greatly decreased as plant density increased.The yield at 6.5 plant/m2 was the highest.
In order to reduce the pressure of water resources in the basin,adjustment and optimized agriculture planting structure,development grass livestock industry,after two years'sowing seeds by stages experiment,growth and development characteristics,yield forming laws and adaptability to climate of the two kinds of grass,Medicago sative and Astragalus adsurgens under the condition of artificial planting were determined.The result indicated:(1) The dry matter weight of Medicago sative and Astragalus adsurgens of the second year(2004) in the spring sowing land was higher than in the autumn sowing land,increasing the yield by 3 505.0 kg and 2 314.5 kg per hm2 respectively.(2) The increase of grass dry weight and plant height sowed in spring and in autumn was basically isochronous and the growth key time all appeared in budding and flowering period.In this period,dry matter accumulation quantity of Medicago sative and Astragalus adsurgens was respectively 70% and 63% of the total.(3) The least value,which was showed by the ratio between dryness and blueness of two kinds of grass growing on the second year(2004) in spring sowing land and autumn sowing land,appeared in flowering period which was also the best time to reap pasture.(4) The grass underground biomass in spring sowing land was better than in autumn sowing land.The grass in spring sowing land was helpful for root system to grow.(5) It was the fitting temperature for pasture to return green that means daily temperature steadily passed 5℃ in spring.Under the condition of irrigation,pasture dry weight was positively associated with heat factor and the productivity was mainly dependent on heat and the requirement of autumn sowing land and late maturing variety for heat condition was more urgent.
In vehicle monitoring, recognizing graphic vehicle identification number (VIN) on the car frame is a particularly important step. While text recognition methods have made great progress, automatic graphic vehicle VIN recognition is still challenging. In VIN images, the VIN text is engraved on the car frame, with complex background and arbitrary orientation, which make it extremely difficult for recognition. We propose an efficient framework for recognizing rotational VIN. First, combining lightweight convolutional neural network and per-pixel segmentation in the output layer, we achieve fast and excellent VIN detection. Second, we take the VIN recognition task as a sequential position-dependent classification problem. By attaching sequential classifiers, we predict VIN text without character segmentation. Finally, we introduce a VIN dataset, which contains 2000 raw rotational VIN images and 90,000 horizontal VIN images for validating our framework. Experiments results show that the framework we proposed achieves good performance in VIN detection and recognition. By automatically identifying the VIN, we can quickly confirm the vehicle's identity and help vehicle monitoring and tracking.
Abstract The “quality project” is an objective need to promote the comprehensive and coordinated development of higher education and cultivate high-quality talents. Based on the fuzzy clustering maximum tree algorithm, this paper first generates the maximum tree for a similar matrix and then clusters the maximum tree to find the clustering results. Then the idea of fuzzy clustering maximum tree algorithm is introduced into the management of “quality engineering” in Heilongjiang province colleges and universities, and the necessity of project management in the management of “quality engineering” in colleges and universities is analyzed, so that the knowledge related to project management can be applied in the construction of “quality engineering” in colleges and universities. “The results show that project quality management has a lot of advantages and disadvantages. The results show that there are problems such as weak awareness of quality management among all staff, unclear construction objectives and quality construction standards, and a lack of effective project monitoring and evaluation mechanism in project quality management. During the experimental period, the groundwater level was 0.4~0.7m below the surface, and the global risk level decreased from 5 to 3 along with the deep excavation of the foundation pit, and the risk probability of the foundation pit project became larger with time and finally stabilized at 60%. The research of this paper is helpful to improve the project management level of “quality engineering” in universities and promote the transformation of “quality engineering” results, which has certain practicality and guidance.
At present, most scene text recognition methods achieve good performance by training models on many synthetic data. However, many data lead to huge storage space and large amount of calculation. And there is a gap between synthetic and real data. To solve these problems, we use a few real data to train a novel proposed model named spatial attention contrastive network (SAC-Net). The SAC-Net consists of a background suppression network (BSNet), a feature encoder, an attention decoder (ADEer), and a feature contrastive network (FCNet). The BSNet based on U-Net is used to reduce the interference of background. For relatively low prediction accuracy brought by connectionist temporal classification, we design an ADEer to improve performance by using convolutional attention mechanism. Based on data augmentation, we design a FCNet which belongs to contrastive learning. Finally, our SAC-Net is almost equivalent to the state-of-the-art model trained on a few real data for word accuracy on six benchmark test datasets.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
