基于数字摄像的测量观测系统设计

在大气能见度观测中，相关研究多集中于白天观测，夜间观测的研究比较少见，而对于昼夜连续观测的研究更是鲜有报道。目前，国内外尚没有完全按照能见度定义研制的大气能见度自动观测仪出现。针对这一问题，基于CCD数字摄像技术，模拟人工观测光学原理，提出了一种能够昼夜进行大气能见度连续观测的方法。该方法基于对目标物获得能见度计算公式，并针对昼夜模式下的系统参数进行修正，能够有效消除外在环境因素和相机自身系统内部因素引发的观测误差。为验证算法的有效性，针对天空遮挡、半遮挡和空旷等不同情况（不同情况，会影响观测环境的亮度，以此测试系统对杂散光的抵抗能力，以及对不同环境的适应能力），搭建了基于该方法的3套原理样机。利用搭建的数字摄像能见度仪（DPVS）在北京地区进行了实际的大气能见度分钟级观测。观测实验表明，基于该方法构建的观测系统，不但具有较宽的观测范围，而且能够有效适应各类复杂天气情况，在雨雪霾等不同天气下，均有较好的观测效果，且无论是在能见度变化较快或者较为缓慢的情况下，DPVS系统均能够进行快速正确的响应。而透过DPVS与散射仪和透射仪这两种标准观测仪器观测结果的对比分析发现，DPVS与前两者的观测结果具有较高的相关性：0.973 1，且观测性能相当，其平均相对误差在-1.54%左右，均方根相对误差在8.82%左右。而本文算法出现的最大相对误差为-14.11%。世界气象组织MOR规定：能见度仪在满量程范围内，其最大相对误差小于20%，即认为是达到标准的能见度仪，可以投入实际观测使用。基于该算法研发的DPVS系统符合观测标准，能够投入实际使用，且DPVS系统的观测成本远比散射仪和透射仪更低，具有较好的前景和应用价值。

Design of Measurement and Observation System Based on Digital Camera

In the atmospheric visibility observation, most of the related research focus on daytime observation, while the researches on nighttime observation are rare, and the research on the continuous observation of day and night are rarely reported. At present, there is no automatic atmospheric visibility observation instrument developed by the definition of visibility at home and abroad. In order to solve this problem, based on CCD digital camera technology, the principle of artificial observation optics is simulated, and a method of continuous observation of atmospheric visibility is proposed. This method is based on the visibility calculation formula of the target. It modifies the system parameters in the day and night mode, which can effectively eliminate the observation error caused by the external environmental factors and the internal factors of the camera system. In order to verify the effectiveness of the algorithm, three sets of principle prototypes based on the method are built for different situations such as sky occlusion, semi occlusion and open space (different situations will affect the brightness of the observation environment, to test the system’s resistance to stray light and adaptability to different environments). They are using the built digital photography visibility system (DPVS) to observe the actual atmospheric visibility at a minute level in the Beijing area. Observation experiments show that the observation system based on this method has a wide observation range and can effectively adapt to all kinds of complex weather conditions. It has a good observation effect in different weather conditions such as rain, snow and haze, and the DPVS system can respond quickly and correctly no matter the visibility changes rapidly or slowly. Through the comparative analysis of the observation results of DPVS, scatterometer and transmission instrument, it is found that there is a high correlation between DPVS and the observation results of the former two instruments: 0.973 1, and the observation performance is similar, with the average relative error of -1.54% and the root mean relative square error of 8.82%. The maximum relative error of this algorithm is -14.11%. According to the World Meteorological Organization (WMO) MOR, if the maximum relative error of the visibility meter is less than 20% within the full range, it is considered a standard visibility meter and can be used in actual observation. The DPVS system based on this algorithm meets the observation standard and can be practical. Moreover, the observation cost of the DPVS system is much lower than that of the scatterometer and transmission instrument, so it has a good prospect and application value.