This study analyzes the measurement errors of three dimensional coordinates of binocular stereo vision for tomatoes based on three stereo matching methods, centroid-based matching, area-based matching, and combination matching to improve the localization accuracy of the binocular stereo vision system of tomato harvesting robots. Centroid-based matching was realized through the matching of the feature points of centroids of tomato regions. Area-based matching was realized based on the gray similarity between two neighborhoods of two pixels to be matched in stereo images. Combination matching was realized using the rough disparity acquired through centroid-based matching as the center of the dynamic disparity range which was used in area-based matching. After stereo matching, three dimensional coordinates of tomatoes were acquired using the triangle range finding principle. Test results based on 225 stereo images captured at the distances from 300 to 1000 mm of 3 tomatoes showed that the measurement errors of x coordinates were small, and can meet the need of harvesting robots. However, the measurement biases of y coordinates and depth values were large, and the measurement variation of depth values was also large. Therefore, the measurement biases of y coordinates and depth values, and the measurement variation of depth values should be corrected in the future researches.
Proc. SPIE. 7658, 5th International Symposium on Advanced Optical Manufacturing and Testing Technologies: Optoelectronic Materials and Devices for Detector, Imager, Display, and Energy Conversion Technology
Accurate harvesting depends on the order of the accuracy of 3D location for harvesting robot. The precision of location
is lower when the distance between fruit and camera is larger than 0.8 m for the method based on binocular stereo vision.
This is a big problem. In order to improve the precision of depth measurement for ripe tomato, two stereo matching
methods were analyzed comparatively which were centroid-based matching and area-based matching. Their
performances in depth measurement were also compared. Experiments showed that the relationship between distance and
measurement was linear. Then, models of unitary linear regression (ULR) were used to improve the results of depth
measurement. After correction by these models, the depth errors were in a range of -28 mm to 25 mm for centroid-based
matching method and -8 mm to 15 mm for area-based matching method at a distance of 0.6 m to 1.15 m. It can be
concluded that costs of computation can be decreased with the promise of good precision when the parallax of centroid
which is acquired through centroid-based matching method is used to set the range of parallax for area-based matching
As a measurement system, the most essential variable of a machine vision measurement system is effective resolution. So far the analysis and evaluation of effective resolution of measurement systems are qualitative. In this paper, a quantitative evaluation method is presented bases on analysis of variance (ANOVA) and correlation analysis. Using this method, effective resolution of a machine vision measurement system can be calculated fleetly. An online evaluation system which bases on this method is introduced.
Based on optical graduation technology, video graduation technology is proposed. The generation of video vertical line, horizontal line and crossing line is introduced. The principle and precision of video graduation and aiming are discussed. At last, an application example of the technology that used in the steel tape fixed length cutting system is presented.
In this paper a new method based on the Position Sensitive Device (PSD) and triangulation for measuring turnplate's run-out error of induction meters was introduced. It lays emphasis on analyzing the working process of the apparatus and the measurement theory and discusses the factors that affect the measuring definition. The design of hardware and software of the apparatus is given and the method of position and calibration of the apparatus is presented.
This paper illustrates the principle and application of dynamic collimation technology based on video graduation. The system that uses this technology could generate video graduation lines dynamically. It can also realize position identification and dynamic collimation of measured objects. This technology has been applied in practice with satisfactory results.