谷歌街景：FOV和缩放之间的转换 [英] Google Streetview: Conversion between FOV and Zoom

问题描述

有谁知道我可以使用Google Maps Android API进行转换的公式：

这是从Google Streetview Image API下载的图像，视频76.8（

这是从Google Streetview Image API下载的FOV 45图片（

我得到的最接近的是如果我制作了FOV 26，但这是我猜测，不使用公式 - 下面的图像是在FOV 26：

解决方案

我终于不得不倒退线以获得指数公式。 >我使用了这个网站，并且手动输入了可以很好地协同工作的FOV和缩放值的数据点组合： http://www.had2know.com/academics/regression-calculator-statistics-best-fit.html

我的数据点是：

FOV（在FOV旁边的括号内放大）

<120>（b）
$ b $ （b）50（1）

45（1.2）

37（1.5）

26（2）

69（0.5）

<78> 78 （0.3）

31（1.9）

为了获得更接近的公式，您可以连续添加更多的很好的数据点，但我没有时间 - 任何想要完善公式的人都可以尝试。

我得到下面的公式其中Y是FOV，X是缩放。

Y = 103.7587（0.5051 ^ X）

相关= -0.9882

Does anyone know of a formula that I can use to convert from Zoom in Google Maps Android API: https://developers.google.com/maps/documentation/android-api/streetview to the FOV used in the Google StreetView Image API: https://developers.google.com/maps/documentation/streetview/intro?

I found a couple of old formulas here:

1. https://groups.google.com/forum/#!topic/google-maps-js-api-v3/uqKfg0ZBhWc
2. https://groups.google.com/forum/#!msg/google-maps-image-apis/O_Odb0A7_0c/Q74cHCoRuscJ

I then put them to a test whereby the FOV will be calculated when I zoom the camera on my android phone:

FOV1: 3.9018*Math.pow(streetViewPanoramaCamera.zoom,2) - 42.432*streetViewPanoramaCamera.zoom + 123

FOV2: Math.abs(streetViewPanoramaCamera.zoom/5*108-120

FOV3: 180/Math.pow(2,streetViewPanoramaCamera.zoom)

@Override
public void onStreetViewPanoramaCameraChange(StreetViewPanoramaCamera streetViewPanoramaCamera) {

    Log.e("Pano", "Bearing: " + streetViewPanoramaCamera.bearing + " Tilt: " + streetViewPanoramaCamera.tilt +
            " Zoom: " + streetViewPanoramaCamera.zoom +
    " FOV1: "+ String.valueOf(3.9018*Math.pow(streetViewPanoramaCamera.zoom,2) - 42.432*streetViewPanoramaCamera.zoom + 123) +" FOV2: "+
            Math.abs(streetViewPanoramaCamera.zoom/5*108-120) +" FOV3: "+ 180/Math.pow(2,streetViewPanoramaCamera.zoom) ) ;

}

At a zoom of 0.0 on android, we have the following FOV values returned:

09-27 15:53:52.322 E/Pano﹕ Bearing: 228.28955 Tilt: 14.516191 Zoom: 0.0 FOV1: 123.0 FOV2: 120.0 FOV3: 180.0

Since FOV has a max of 120, FOV2's formula seemed promising at first but when I zoomed at 2 times, it gave me a value of 76.8 which is far off from the actual:

09-27 16:01:48.235 E/Pano﹕ Bearing: 223.11241 Tilt: 1.852709 Zoom: 2.0 FOV1: 53.7432 FOV2: 76.8 FOV3: 45.0

This is the image on my phone after the 2 times zoom:

This is the image downloaded from Google Streetview Image API at FOV 76.8 (https://maps.googleapis.com/maps/api/streetview?size=600x300&location=-33.87365,151.20689&heading=223.11241&pitch=1.852709&fov=76.8&key=APIKEY):

This is the image downloaded from Google Streetview Image API at FOV 45 (https://maps.googleapis.com/maps/api/streetview?size=600x300&location=-33.87365,151.20689&heading=223.11241&pitch=1.852709&fov=45&key=APIKEY):

The closest I have gotten is if I made the FOV 26 but that is me guessing and not using the formulas - image below is at FOV 26:

解决方案

I finally had to regress the line to get the exponential formula out.

I used this website and manually typed in combinations of data points of both FOV and zoom values that worked well together: http://www.had2know.com/academics/regression-calculator-statistics-best-fit.html

My data points were:

FOV (ZOOM in brackets next to FOV)

120 (0)

50 (1)

45 (1.2)

37 (1.5)

26 (2)

69 (0.5)

78 (0.3)

31 (1.9)

To get a closer formula, you could continuously add on more "good-fit" data points but I just didn't have the time - anyone else who would like to perfect the formula can attempt it though.

I got the formula below where Y is the FOV and X is the zoom.

Y = 103.7587(0.5051^X)

correlation = -0.9882

这篇关于谷歌街景：FOV和缩放之间的转换的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！