(转)Android屏幕截图实现 (adbd部分)

/android2.1/system/core/adb/framebuffer_service.c  这个里面是DDMS 里面截图通信adbd进程的代码   DDMS那边接到的数据是用JAVA处理的  ！   framebuffer_service.c代码如下：

 

 

/*
 * Copyright (C) 2007 The Android Open Source Project
 *
 * Licensed under the Apache License, Version 2.0 (the “License”);
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an “AS IS” BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#include <stdlib.h>
#include <stdio.h>
#include <unistd.h>
#include <string.h>
#include <fcntl.h>

#include “fdevent.h”
#include “adb.h”

#include <linux/fb.h>
#include <sys/ioctl.h>
#include <sys/mman.h>

/* TODO:
** – sync with vsync to avoid tearing
*/
/* This version number defines the format of the fbinfo struct.
   It must match versioning in ddms where this data is consumed. */
#define DDMS_RAWIMAGE_VERSION 1
struct fbinfo {
    unsigned int version;
    unsigned int bpp;
    unsigned int size;
    unsigned int width;
    unsigned int height;
    unsigned int red_offset;
    unsigned int red_length;
    unsigned int blue_offset;
    unsigned int blue_length;
    unsigned int green_offset;
    unsigned int green_length;
    unsigned int alpha_offset;
    unsigned int alpha_length;
} __attribute__((packed));

void framebuffer_service(int fd, void *cookie)
{
    struct fb_var_screeninfo vinfo;
    int fb, offset;
    char x[256];

    struct fbinfo fbinfo;
    unsigned i, bytespp;

    fb = open(“/dev/graphics/fb0″, O_RDONLY);
    if(fb < 0) goto done;

    if(ioctl(fb, FBIOGET_VSCREENINFO, &vinfo) < 0) goto done;
    fcntl(fb, F_SETFD, FD_CLOEXEC);

    bytespp = vinfo.bits_per_pixel / 8;

    fbinfo.version = DDMS_RAWIMAGE_VERSION;
    fbinfo.bpp = vinfo.bits_per_pixel;
    fbinfo.size = vinfo.xres * vinfo.yres * bytespp;
    fbinfo.width = vinfo.xres;
    fbinfo.height = vinfo.yres;
    fbinfo.red_offset = vinfo.red.offset;
    fbinfo.red_length = vinfo.red.length;
    fbinfo.green_offset = vinfo.green.offset;
    fbinfo.green_length = vinfo.green.length;
    fbinfo.blue_offset = vinfo.blue.offset;
    fbinfo.blue_length = vinfo.blue.length;
    fbinfo.alpha_offset = vinfo.transp.offset;
    fbinfo.alpha_length = vinfo.transp.length;

    /* HACK: for several of our 3d cores a specific alignment
     * is required so the start of the fb may not be an integer number of lines
     * from the base.  As a result we are storing the additional offset in
     * xoffset. This is not the correct usage for xoffset, it should be added
     * to each line, not just once at the beginning */
    offset = vinfo.xoffset * bytespp;

    offset += vinfo.xres * vinfo.yoffset * bytespp;

    if(writex(fd, &fbinfo, sizeof(fbinfo))) goto done;

    lseek(fb, offset, SEEK_SET);
    for(i = 0; i < fbinfo.size; i += 256) {
      if(readx(fb, &x, 256)) goto done;
      if(writex(fd, &x, 256)) goto done;
    }

    if(readx(fb, &x, fbinfo.size % 256)) goto done;
    if(writex(fd, &x, fbinfo.size % 256)) goto done;

done:
    if(fb >= 0) close(fb);
    close(fd);
}

       

用framebuffer截屏，就是把当前framebuffer的内存dump成bitmap或者压缩成其他图片格式.

 

 

在ubuntu下使用命令截屏：

 

Android G1手机的色深是16bit, 即R/G/B=5/6/6. 故需要一些特殊处理才能获得其屏幕图像。我们可以adb登录进入Android手机，先dump出/dev/fb0的内容，然后将dump出的文件拷贝到PC（ubuntu）上用convert命令进行处理。 1. 在Android手机上获取framebuffer内容： # cat /dev/graphics/fb0 > /mnt/sdcard/fb0 2. 把文件拷贝到PC上的/tmp目录： $ adb pull /mnt/sdcard/fb0 /tmp/fb0 3. 从/tmp/fb0文件中取出第一屏的内容（一般会有2到3屏的内容）： $ dd bs=307200 count=1 if=/tmp/fb0 of=/tmp/screenshot1.xmp 这里，手机屏幕分辨率是320*480; 颜色深度为16bit (R/G/B=5/6/5)，即2字节; 故有 320*480*2=307200 或在shell下执行： $ echo $((320*480*2)) 这里得到的图像是pixmap格式的. 4. 这时取到的截屏数据要转成每个颜色值为8bit的raw图像数据: 原理如下： int main(int argc, char *argv[]) { unsigned short in; // 16bit unsigned char out[3]; // 8bit*3 while (read(0, &in, 2) == 2) { out[2] = (in & 0x1f) << 3; out[1] = ((in >> 5) & 0x3f) << 2; out[0] = ((in >> 11) & 0x1f) << 3; write(1, out, 3); } return 0; } 将以上代码编译成执行文件/tmp/565to88后，再执行： $ /tmp/565to888 /tmp/screenshot888.xmp 这将把颜色转化为24bit色深。 5.最后把24bit色深的图像转换成PNG格式： $ /usr/bin/convert -depth 8 -size 320×480 rgb:screenshot888.xmp screenshot.png 现在可以用以下命令查看截屏(screenshot.png文件)啦： $ /usr/bin/display screenshot.png

转自：http://blog.csdn.net/yuleslie/article/details/7003431