Food recognition using neural network classifier and multiple hypotheses image segmentation

ABSTRACT

This paper proposes the multiple-hypotheses image segmentation and feed-forward neural network classifier for food recognition to improve the performance. Initially, the food or meal image is given as input. Then, the segmentation is applied to identify the regions, where a particular food item is located using salient region detection, multi-scale segmentation, and fast rejection. Then, the features of every food item are extracted by the global feature and local feature extraction. After the features are obtained, the classification is performed for each segmented region using a feed-forward neural network model. Finally, the calorie value is computed with the aid of (i) food volume and (ii) calorie and nutrition measure based on mass value. The experimental results and performance evaluation are validated. The outcome of the proposed method attains 0.947 for Macro Average Accuracy (MAA) and 0.959 for Standard Accuracy (SA), which provides better classification performance.     

测试(试井)成果影响因素浅析

随着测试技术的发展、测试质量及资料解释水平的提高，测试（试井）成果的可靠程度越来越受到重视。影响解释成果的环节和因素很多，从客观和主观两方面分析其主要因素，找出症结所在，从而完善现场环节，完善资料解释过程，提高测试技术及解释水平。     

基于承包商预期收益不变的不平衡报价模型

在招标实践中,通常采用最低价中标原则,使得投标人不得不压低其投标价格。基于承包商预期收益不变的不平衡报价模型,是在“工程量清单报价”的条件下,考虑到施工中工程量的变化及施工顺序的影响,体现了资金的时间价值,在保持承包商预期收益不变的前提下,通过调整各分项工程的单价,来降低工程报价,以增加中标机会而建立的。所建立的模型简单,易于操作,并配以具体的工程实例,通过计算机求得其最优解,可供承包商投标报价时参考。     

On keys' meanings and modes: the impact of different key solutions on children's efficiency using a mobile phone

The present study investigates the impact of different key solutions of mobile phones on users' effectiveness and efficiency using the devices. In the first experiment, 36 children (9 - 14 years) and in the second experiment 45 young adults (19 - 33 years) completed four common phone tasks twice consecutively on three simulated phones that had identical menus, but different key solutions. An approach was undertaken to quantify the complexity of keys in three models, incorporating different factors contributing to the keys' complexity (number of key options, number of modes and number of modes with a semantically dissimilar meaning), in order to predict users' performance decrements. As a further main factor, the degree of the users' locus of control (LOC) was measured and interactions with performance outcomes were studied. As dependent measures, the number of inefficient keystrokes, the number of tasks solved and the processing time were determined. Results showed a significant effect of control key solutions on users' efficiency and effectiveness for both children and young adults. Moreover, children's LOC values significantly interacted with performance: children with low LOC values showed the lowest performance and no learnability, especially when using keys with a high complexity. From the three factors contributing to the complexity of keys, keys exerting different functions with semantically inconsistent meanings had the worst effect on performance. It is concluded that in mobile user interface design keys with semantically inconsistent meanings should be generally avoided.     

关于中值定理的几点应用

本文作者给出了在两种不同类型的证明题中应用中值定理的方法.这些方法简单、实用.     

人工神经网络预测煤的发热量

应用人工神经网络处理部分广东地区煤质分析数据,建立网络模型,预测出煤的发热量,经实测值与预测值比较,在一定的煤种范围内预测值与实测值比较接近。     

试油井产量折算新方法

为了使试油折算产量的条件尽可能的相同,增强试油产量的可比性,消除地面计量的人为误差,研究了试油求产时井底流压的变化规律,认为采用油柱在油管内上升的流压变化速率来折算产量更趋于合理,并能最大限度地接近地层真实产量.该方法从根本上消除了因试油周期选定不合理、地面计量误差,以及试油工艺不同、储层条件差异带来的试油产量误差.     

天然水资源价值评估的能值方法及应用

本文从水资源的自然属性出发，提出了采用能值(Emergy)理论与方法研究天然水资源价值的新思路。论述了环境与自然资源定价的能值价值理论；根据水文循环与天然水资源基本特性，给出了水资源生成与能量转换的概念模型；从水的化学能角度提出了天然水资源价值的能值评估方法，包括理论方法与简化方法；并以我国南方某沿海县域为例进行了实际应用。实例表明，基于能值价值理论的天然水资源价值定量评估方法具有可行性。     

有限元技术在数值试井中的应用

分析了试井技术的发展趋势，对有限元分析技术及其发展作了较为全面的介绍，重点阐述了利用有限元技术进行三维建模的方法和步骤，尤其是对复杂的非对称边界和非均质油藏的建模方法和技巧进行了深入探讨。利用数值试井技术，石油工程师可以方便准确地对复杂油藏和边界进行建模，试井的模拟过程和油藏任意位置的压力瞬间变化都可以通过图形和颜色的变化显示出来，解释的精度和速度都得到了大大地提高。     

基于系统矩阵有序实Schur分解的模型降阶

对复杂高阶系统进行模型简化，为研究与设计系统提供了方便的条件。对系统模型的降阶是控制系统设计与仿真工作者的一个重要研究课题，对高阶复杂系统的分析、设计、仿真等具有重要的理论意义和工程实用价值。针对目前比较成熟的基于系统矩阵的有序实Schur分解方法，提出了采用时矩输出拟合来改善降阶系统输出响应逼近程度的改进方法，从而使简化模型具有输出误差更小、计算简便等优点，通过给出的实例仿真证实了这一点。