前言

镜头阴影校正（Lens Shading Correction）是为了解决由于lens的光学特性，由于镜头对于光学折射不均匀导致的镜头周围出现阴影的情况。
由于Lens的光学特性，Sensor影像区的边缘区域接收的光强比中心小，所造成的中心和四角亮度不一致的现象。镜头本身就是一个凸透镜，由于凸透镜原理，中心的感光必然比周边多。如图所示：

实际成像效果如图：

中间区域发亮，四角发黑

海康工业相机SDK LSC算法矫正接口

SDK下载获取

海康机器视觉sdk简介，下载链接：机器视觉下载中心，下载安装MVS即可

示例程序demo，参考“C:\Program Files (x86)\MVS\Development\Samples\VC\VS\SimpleSamples*LensShadingCorrection*”

代码流程简介

LSC矫正主要分为标定文件生成、lsc矫正使用

LSC标定文件生成

核心思想是在取流接口中，或者到一张待标定的图片，调用标定接口进行标定文件生成
如下代码，再回调函数中，调用MV_CC_LSCCalib接口，生成LSCCalib.bin标定文件

void __stdcall ImageCallBackEx(unsigned char * pData, MV_FRAME_OUT_INFO_EX* pFrameInfo, void* pUser)
{printf("Get One Frame: Width[%d], Height[%d], nFrameNum[%d]\n", pFrameInfo->nWidth, pFrameInfo->nHeight, pFrameInfo->nFrameNum);int nRet = MV_OK;//判断是否需要标定if (true == g_IsNeedCalib){// LSC标定MV_CC_LSC_CALIB_PARAM stLSCCalib = {0};stLSCCalib.nWidth = pFrameInfo->nWidth;stLSCCalib.nHeight = pFrameInfo->nHeight;stLSCCalib.enPixelType = pFrameInfo->enPixelType;stLSCCalib.pSrcBuf = pData;stLSCCalib.nSrcBufLen = pFrameInfo->nFrameLen;if (g_pCalibBuf == NULL || g_nCalibBufSize < (pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short))){if (g_pCalibBuf){free(g_pCalibBuf);g_pCalibBuf = NULL;g_nCalibBufSize = 0;}g_pCalibBuf = (unsigned char *)malloc(pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short));if (g_pCalibBuf == NULL){printf("malloc pCalibBuf fail !\n");return;}g_nCalibBufSize = pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short);}stLSCCalib.pCalibBuf = g_pCalibBuf;stLSCCalib.nCalibBufSize = g_nCalibBufSize;stLSCCalib.nSecNumW = 689;stLSCCalib.nSecNumH = 249;stLSCCalib.nPadCoef = 2;stLSCCalib.nCalibMethod = 2;stLSCCalib.nTargetGray = 100;//同一个相机，在场景、算法参数和图像参数都不变情况下，理论上只需要进行一次标定，可将标定表保存下来。//不同相机图片标定出来的标定表不可复用，当场景改变或算法参数改变或图像参数改变时，需要重新标定。nRet = MV_CC_LSCCalib(pUser, &stLSCCalib);if (MV_OK != nRet){printf("LSC Calib fail! nRet [0x%x]\n", nRet);return;}//保存标定表到本地FILE* fp_out = fopen("./LSCCalib.bin", "wb");if (NULL == fp_out){return ;}fwrite(stLSCCalib.pCalibBuf, 1, stLSCCalib.nCalibBufLen, fp_out);fclose(fp_out);g_IsNeedCalib = false;}
}

MV_CC_LSCCalib接口中，需要传入算法参数，来看一下这个结构体MV_CC_LSC_CALIB_PARAM的构成
除了常规的图像宽、高、像素格式之外的参数，部分算法参数如下解释：

参数	解释	建议值
nSecNumW	宽度分快数，nSecNumW越大，矫正后一致性越好，但此值越大，会导致部分缺陷类检测缺陷失真；nSecNumW<widthmax	图像最大宽度除以4
nSecNumH	高度分快数，nSecNumH越大，矫正后一致性越好，但此值越大，会导致部分缺陷类检测缺陷失真；nSecNumH<hightmax	图像最大高度除以4
nPadCoef	边缘填充系数，范围1-5	2
nCalibMethod	标定方式[0-2]，提供三种标定方法1：中心为亮度基准，取图像中心区域灰度值为基准进行矫正；2：最亮区域；3：目标亮度，像数格式8bit时，nTargetGray值范围可设置为[0-255];像数格式10bit时，nTargetGray值范围可设置为[0-1023];像数格式12bit时，nTargetGray值范围可设置为[0-4095]	0
nTargetGray	nCalibMethod取值为2时有效	nCalibMethod为2时，图像中心区域平均亮度

typedef struct _MV_CC_LSC_CALIB_PARAM_T_
{unsigned int            nWidth;                                 ///< [IN]  \~chinese 图像宽度[16,65535]     \~english Image Widthunsigned int            nHeight;                                ///< [IN]  \~chinese 图像高度[16-65535]     \~english Image Heightenum MvGvspPixelType    enPixelType;                            ///< [IN]  \~chinese 像素格式               \~english Pixel formatunsigned char*          pSrcBuf;                                ///< [IN]  \~chinese 输入数据缓存           \~english Input data bufferunsigned int            nSrcBufLen;                             ///< [IN]  \~chinese 输入数据长度           \~english Input data lengthunsigned char*          pCalibBuf;                              ///< [OUT] \~chinese 输出标定表缓存         \~english Output calib bufferunsigned int            nCalibBufSize;                          ///< [IN]  \~chinese 提供的标定表缓冲大小(nWidth*nHeight*sizeof(unsigned short))    \~english Provided output buffer sizeunsigned int            nCalibBufLen;                           ///< [OUT] \~chinese 输出标定表缓存长度     \~english Output calib buffer lengthunsigned int            nSecNumW;                               ///< [IN]  \~chinese 宽度分块数             \~english Width Sec numunsigned int            nSecNumH;                               ///< [IN]  \~chinese 高度分块数             \~english Height Sec numunsigned int            nPadCoef;                               ///< [IN]  \~chinese 边缘填充系数[1,5]      \~english Pad Coef[1,5]unsigned int            nCalibMethod;                           ///< [IN]  \~chinese 标定方式(0-中心为基准;1-最亮区域为基准;2-目标亮度为基准) \~english Calib methodunsigned int            nTargetGray;                            ///< [IN]  \~chinese 目标亮度(标定方式为2时有效)    \~english Target Gray///< \~chinese 8位,范围：[0,255]            \~english 8bit,range:[0,255]///< \~chinese 10位,范围：[0,1023]          \~english 10bit,range:[0,1023]///< \~chinese 12位,范围：[0,4095]          \~english 12bit,range:[0,4095]unsigned int            nRes[8];                                ///<       \~chinese 预留                   \~english Reserved
}MV_CC_LSC_CALIB_PARAM;

标定过程，需要特定场景的，按照如下步骤进行
1、相机对准均匀场景（均匀光源、白纸、白板等），通过调节曝光、增益等，使得图像灰度在120-160之间（镜头光圈保持与实际使用场景一致）

2、运行标定程序，生成标定文件；可直接运行LensShadingCorrection示例demo，exe路径下会有当前标定文件生成
3、拍摄实际物体，运行矫正程序，观察效果；可直接运行LensShadingCorrection示例demo观察
如果标定矫正效果不佳，那么就需要重新调整标定场景，重新生成标定文件
同理，如遇到多种光源切换场景，可以生成不同的标定文件，进行动态切换矫正

LSC矫正

拿到标定文件后，矫正过程相对简单，获取一张图像后，再调用MV_CC_LSCCorrect接口即可完成（每一张图片都需要调用，一张一张矫正）

void __stdcall ImageCallBackEx(unsigned char * pData, MV_FRAME_OUT_INFO_EX* pFrameInfo, void* pUser)
{printf("Get One Frame: Width[%d], Height[%d], nFrameNum[%d]\n", pFrameInfo->nWidth, pFrameInfo->nHeight, pFrameInfo->nFrameNum);int nRet = MV_OK;// LSC校正if (g_pDstData == NULL || g_nDstDataSize < pFrameInfo->nFrameLen){if (g_pDstData){free(g_pDstData);g_pDstData = NULL;g_nDstDataSize = 0;}g_pDstData = (unsigned char *)malloc(pFrameInfo->nFrameLen);if (g_pDstData == NULL){printf("malloc pDstData fail !\n");return;}g_nDstDataSize = pFrameInfo->nFrameLen;}MV_CC_LSC_CORRECT_PARAM stLSCCorrectParam = {0};stLSCCorrectParam.nWidth = pFrameInfo->nWidth;stLSCCorrectParam.nHeight = pFrameInfo->nHeight;stLSCCorrectParam.enPixelType = pFrameInfo->enPixelType;stLSCCorrectParam.pSrcBuf = pData;stLSCCorrectParam.nSrcBufLen = pFrameInfo->nFrameLen;stLSCCorrectParam.pDstBuf = g_pDstData;stLSCCorrectParam.nDstBufSize = g_nDstDataSize;stLSCCorrectParam.pCalibBuf = g_pCalibBuf;stLSCCorrectParam.nCalibBufLen = g_nCalibBufSize;nRet = MV_CC_LSCCorrect(pUser, &stLSCCorrectParam);if (MV_OK != nRet){printf("LSC Correct fail! nRet [0x%x]\n", nRet);return;}//矫正完成，其他图像处理
}

矫正耗时测试

测试1：不同分辨率图像，不同版本算法，耗时情况

分辨率	x64	x32
2448*2048	4.3ms	5.6ms
3840*2748	8.6ms	11.89ms
5472*3648	16.45ms	22.14ms

算法耗时取决于图像分辨率大小，分辨率越大，耗时越大
测试2：不同算法参数，矫正模块耗时情况

分辨率	nCalibMethod	耗时	分块数	耗时	边缘填充系数	耗时
2448*2048	0：中心基准	4.25ms	16*16	4.28ms	1	4.26ms
2448*2048	1：最亮基准	4.25ms	600*400	4.30ms	3	4.20ms
2448*2048	2：目标基准	4.28ms	2448*2048	4.25ms	5	4.26ms

图像效果对比

测试参数：

参数	Value
分辨率	2448*2048
nSecNumW	2448
nSecNumH	2048
nPadCoef	2
nCalibMethod	0

LensShadingCorrection完整示例程序

#include <stdio.h>
#include <Windows.h>
#include <conio.h>
#include <io.h>
#include "MvCameraControl.h"
// ch:等待按键输入 | en:Wait for key press
void WaitForKeyPress(void)
{while(!_kbhit()){Sleep(10);}_getch();
}
bool PrintDeviceInfo(MV_CC_DEVICE_INFO* pstMVDevInfo)
{if (NULL == pstMVDevInfo){printf("The Pointer of pstMVDevInfo is NULL!\n");return false;}if (pstMVDevInfo->nTLayerType == MV_GIGE_DEVICE){int nIp1 = ((pstMVDevInfo->SpecialInfo.stGigEInfo.nCurrentIp & 0xff000000) >> 24);int nIp2 = ((pstMVDevInfo->SpecialInfo.stGigEInfo.nCurrentIp & 0x00ff0000) >> 16);int nIp3 = ((pstMVDevInfo->SpecialInfo.stGigEInfo.nCurrentIp & 0x0000ff00) >> 8);int nIp4 = (pstMVDevInfo->SpecialInfo.stGigEInfo.nCurrentIp & 0x000000ff);// ch:打印当前相机ip和用户自定义名字 | en:print current ip and user defined nameprintf("CurrentIp: %d.%d.%d.%d\n" , nIp1, nIp2, nIp3, nIp4);printf("UserDefinedName: %s\n\n" , pstMVDevInfo->SpecialInfo.stGigEInfo.chUserDefinedName);}else if (pstMVDevInfo->nTLayerType == MV_USB_DEVICE){printf("UserDefinedName: %s\n", pstMVDevInfo->SpecialInfo.stUsb3VInfo.chUserDefinedName);printf("Serial Number: %s\n", pstMVDevInfo->SpecialInfo.stUsb3VInfo.chSerialNumber);printf("Device Number: %d\n\n", pstMVDevInfo->SpecialInfo.stUsb3VInfo.nDeviceNumber);}else{printf("Not support.\n");}return true;
}
unsigned char * g_pDstData = NULL;
unsigned int g_nDstDataSize = 0;
unsigned char * g_pCalibBuf = NULL;
unsigned int g_nCalibBufSize = 0;
bool g_IsNeedCalib = true;
void __stdcall ImageCallBackEx(unsigned char * pData, MV_FRAME_OUT_INFO_EX* pFrameInfo, void* pUser)
{printf("Get One Frame: Width[%d], Height[%d], nFrameNum[%d]\n", pFrameInfo->nWidth, pFrameInfo->nHeight, pFrameInfo->nFrameNum);int nRet = MV_OK;//判断是否需要标定if (true == g_IsNeedCalib){// LSC标定MV_CC_LSC_CALIB_PARAM stLSCCalib = {0};stLSCCalib.nWidth = pFrameInfo->nWidth;stLSCCalib.nHeight = pFrameInfo->nHeight;stLSCCalib.enPixelType = pFrameInfo->enPixelType;stLSCCalib.pSrcBuf = pData;stLSCCalib.nSrcBufLen = pFrameInfo->nFrameLen;if (g_pCalibBuf == NULL || g_nCalibBufSize < (pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short))){if (g_pCalibBuf){free(g_pCalibBuf);g_pCalibBuf = NULL;g_nCalibBufSize = 0;}g_pCalibBuf = (unsigned char *)malloc(pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short));if (g_pCalibBuf == NULL){printf("malloc pCalibBuf fail !\n");return;}g_nCalibBufSize = pFrameInfo->nWidth*pFrameInfo->nHeight*sizeof(unsigned short);}stLSCCalib.pCalibBuf = g_pCalibBuf;stLSCCalib.nCalibBufSize = g_nCalibBufSize;stLSCCalib.nSecNumW = 689;stLSCCalib.nSecNumH = 249;stLSCCalib.nPadCoef = 2;stLSCCalib.nCalibMethod = 2;stLSCCalib.nTargetGray = 100;//同一个相机，在场景、算法参数和图像参数都不变情况下，理论上只需要进行一次标定，可将标定表保存下来。//不同相机图片标定出来的标定表不可复用，当场景改变或算法参数改变或图像参数改变时，需要重新标定。nRet = MV_CC_LSCCalib(pUser, &stLSCCalib);if (MV_OK != nRet){printf("LSC Calib fail! nRet [0x%x]\n", nRet);return;}//保存标定表到本地FILE* fp_out = fopen("./LSCCalib.bin", "wb");if (NULL == fp_out){return ;}fwrite(stLSCCalib.pCalibBuf, 1, stLSCCalib.nCalibBufLen, fp_out);fclose(fp_out);g_IsNeedCalib = false;}// LSC校正if (g_pDstData == NULL || g_nDstDataSize < pFrameInfo->nFrameLen){if (g_pDstData){free(g_pDstData);g_pDstData = NULL;g_nDstDataSize = 0;}g_pDstData = (unsigned char *)malloc(pFrameInfo->nFrameLen);if (g_pDstData == NULL){printf("malloc pDstData fail !\n");return;}g_nDstDataSize = pFrameInfo->nFrameLen;}MV_CC_LSC_CORRECT_PARAM stLSCCorrectParam = {0};stLSCCorrectParam.nWidth = pFrameInfo->nWidth;stLSCCorrectParam.nHeight = pFrameInfo->nHeight;stLSCCorrectParam.enPixelType = pFrameInfo->enPixelType;stLSCCorrectParam.pSrcBuf = pData;stLSCCorrectParam.nSrcBufLen = pFrameInfo->nFrameLen;stLSCCorrectParam.pDstBuf = g_pDstData;stLSCCorrectParam.nDstBufSize = g_nDstDataSize;stLSCCorrectParam.pCalibBuf = g_pCalibBuf;stLSCCorrectParam.nCalibBufLen = g_nCalibBufSize;nRet = MV_CC_LSCCorrect(pUser, &stLSCCorrectParam);if (MV_OK != nRet){printf("LSC Correct fail! nRet [0x%x]\n", nRet);return;}if (pFrameInfo->nFrameNum < 10){//保存图像到文件MV_SAVE_IMG_TO_FILE_PARAM stSaveFileParam;memset(&stSaveFileParam, 0, sizeof(MV_SAVE_IMG_TO_FILE_PARAM));stSaveFileParam.enImageType = MV_Image_Bmp;stSaveFileParam.enPixelType = pFrameInfo->enPixelType;stSaveFileParam.nWidth      = pFrameInfo->nWidth;stSaveFileParam.nHeight     = pFrameInfo->nHeight;stSaveFileParam.nDataLen    = pFrameInfo->nFrameLen;stSaveFileParam.pData       = pData;sprintf_s(stSaveFileParam.pImagePath, 256, "BeforeImage_w%d_h%d_fn%03d.bmp", stSaveFileParam.nWidth, stSaveFileParam.nHeight, pFrameInfo->nFrameNum);nRet = MV_CC_SaveImageToFile(pUser, &stSaveFileParam);if (MV_OK != nRet){printf("SaveImageToFile failed[%x]!\n", nRet);return;}stSaveFileParam.pData       = g_pDstData;sprintf_s(stSaveFileParam.pImagePath, 256, "AfterImage_w%d_h%d_fn%03d.bmp", stSaveFileParam.nWidth, stSaveFileParam.nHeight, pFrameInfo->nFrameNum);nRet = MV_CC_SaveImageToFile(pUser, &stSaveFileParam);if (MV_OK != nRet){printf("SaveImageToFile failed[%x]!\n", nRet);return;}}
}
int main()
{int nRet = MV_OK;void* handle = NULL;do {// ch:枚举设备 | en:Enum deviceMV_CC_DEVICE_INFO_LIST stDeviceList;memset(&stDeviceList, 0, sizeof(MV_CC_DEVICE_INFO_LIST));nRet = MV_CC_EnumDevices(MV_GIGE_DEVICE | MV_USB_DEVICE, &stDeviceList);if (MV_OK != nRet){printf("Enum Devices fail! nRet [0x%x]\n", nRet);break;}if (stDeviceList.nDeviceNum > 0){for (unsigned int i = 0; i < stDeviceList.nDeviceNum; i++){printf("[device %d]:\n", i);MV_CC_DEVICE_INFO* pDeviceInfo = stDeviceList.pDeviceInfo[i];if (NULL == pDeviceInfo){break;} PrintDeviceInfo(pDeviceInfo);            }  } else{printf("Find No Devices!\n");break;}printf("Please Input camera index(0-%d):", stDeviceList.nDeviceNum-1);unsigned int nIndex = 0;scanf_s("%d", &nIndex);if (nIndex >= stDeviceList.nDeviceNum){printf("Input error!\n");break;}// ch:选择设备并创建句柄 | en:Select device and create handlenRet = MV_CC_CreateHandle(&handle, stDeviceList.pDeviceInfo[nIndex]);if (MV_OK != nRet){printf("Create Handle fail! nRet [0x%x]\n", nRet);break;}// ch:打开设备 | en:Open devicenRet = MV_CC_OpenDevice(handle);if (MV_OK != nRet){printf("Open Device fail! nRet [0x%x]\n", nRet);break;}// ch:探测网络最佳包大小(只对GigE相机有效) | en:Detection network optimal package size(It only works for the GigE camera)if (stDeviceList.pDeviceInfo[nIndex]->nTLayerType == MV_GIGE_DEVICE){int nPacketSize = MV_CC_GetOptimalPacketSize(handle);if (nPacketSize > 0){nRet = MV_CC_SetIntValue(handle,"GevSCPSPacketSize",nPacketSize);if(nRet != MV_OK){printf("Warning: Set Packet Size fail nRet [0x%x]!", nRet);}}else{printf("Warning: Get Packet Size fail nRet [0x%x]!", nPacketSize);}}// ch:设置触发模式为off | eb:Set trigger mode as offnRet = MV_CC_SetEnumValue(handle, "TriggerMode", MV_TRIGGER_MODE_OFF);if (MV_OK != nRet){printf("Set Trigger Mode fail! nRet [0x%x]\n", nRet);break;}//判断是否可以本地导入FILE* fp = fopen("./LSCCalib.bin", "rb");if (fp){int nFileLen = filelength(fileno(fp));if (g_pCalibBuf == NULL || g_nCalibBufSize < nFileLen){if (g_pCalibBuf){free(g_pCalibBuf);g_pCalibBuf = NULL;g_nCalibBufSize = 0;}g_pCalibBuf = (unsigned char *)malloc(nFileLen);if (g_pCalibBuf == NULL){printf("malloc pCalibBuf fail !\n");break;}g_nCalibBufSize = nFileLen;}fread(g_pCalibBuf, 1, g_nCalibBufSize, fp);fclose(fp);g_IsNeedCalib = false;}// ch:注册抓图回调 | en:Register image callbacknRet = MV_CC_RegisterImageCallBackEx(handle, ImageCallBackEx, handle);if (MV_OK != nRet){printf("Register Image CallBack fail! nRet [0x%x]\n", nRet);break;}// ch:开始取流 | en:Start grab imagenRet = MV_CC_StartGrabbing(handle);if (MV_OK != nRet){printf("Start Grabbing fail! nRet [0x%x]\n", nRet);break;}printf("Press a key to stop grabbing.\n");WaitForKeyPress();Sleep(1000);// ch:停止取流 | en:Stop grab imagenRet = MV_CC_StopGrabbing(handle);if (MV_OK != nRet){printf("Stop Grabbing fail! nRet [0x%x]\n", nRet);break;}// ch:关闭设备 | en:Close devicenRet = MV_CC_CloseDevice(handle);if (MV_OK != nRet){printf("Close Device fail! nRet [0x%x]\n", nRet);break;}// ch:销毁句柄 | en:Destroy handlenRet = MV_CC_DestroyHandle(handle);if (MV_OK != nRet){printf("Destroy Handle fail! nRet [0x%x]\n", nRet);break;}} while (0);if (g_pCalibBuf){free(g_pCalibBuf);g_pCalibBuf = NULL;g_nCalibBufSize = 0;}if (g_pDstData){free(g_pDstData);g_pDstData = NULL;g_nDstDataSize = 0;}if (nRet != MV_OK){if (handle != NULL){MV_CC_DestroyHandle(handle);handle = NULL;}}printf("Press a key to exit.\n");WaitForKeyPress();return 0;
}

海康工业相机镜头阴影矫正LSC相关推荐

海康工业网口相机组播功能
海康工业网口相机组播功能 1.组播功能MVS演示 1.1网络拓扑图 1.2MVS功能演示 2.组播功能SDK接口调用实现 3.组播功能使用常见问题什么是组播? 网口相机的组播功能可以实现多个PC对同 ...
海康工业摄像头调用（linux基于python和opencv）
1.下载官网客户端(其中包含SDK) 官方网站海康机器人-机器视觉-下载中心安装deb文件 sudo dpkg -i deb文件名 2.运行客户端 cd /opt/MVS/bin/ ./MVS. ...
工业视觉-西门子-1200与海康相机软件通讯
简述: 本文主要讲解海康工业视觉特征软件VisionMaster 4.0.0与西门子1214C型CPU之间建立通讯的配置. 设备: PC:安装VisionMaster 4.0.0,安装TIA Port ...
C++下OPENCV驱动调用海康GigE工业相机
系列文章目录第一章 Ubuntu22下OpenCV4.6.0+contrib模块编译安装第二章 ubuntu22下C++ kdevelop环境搭建:OpenCV示例第三章 C++下OPENCV驱 ...
海康机器人工业相机常用参数功能设置与获取（持续更新全-C语言）
海康机器人工业相机常用参数功能设置与获取(C语言) 前言相机属性树 1.设备控制/DeviceControl 1.1相机序列号获取/DeviceSerialNumber 1.2相机自定义命名设置与获 ...
python 调用海康工业相机调试后出现被占用问题
问题描述: 在使用海康工业网口相机开发过程中,调试阶段异常结束程序调试时,经常出现相机被占用的异常情况,无论是使用海康的客户端界面连接相机还是使用调试代码连接相机,都没有权限再次打开相机,需要相机断电 ...
iNeuOS工业互联平台，WEB组态（iNeuView）集成rtmp和websocket视频元件，支持海康、大华等摄像头实时显示视频
目录 1. 概述... 1 2. 平台演示... 2 3. 硬件摄像头... 2 4. 视频流协议转换管理... 2 5. 组态视频元件 ...
海康工业相机常用参数修改以及使用OpenCV-Python采集图像
海康工业相机常用参数修改以及使用OpenCV-Python采集图像一.调节相机参数 1.调节曝光时间 2.修改图像尺寸和像素格式 3.保存参数二.Python采集成opencv可识别格式,并用op ...
10 海康视觉平台VisionMaster 上手系列：硬件知识介绍
硬件知识,不属于VisionMaster应用.属于机器视觉应用. 机器视觉硬件: 1,相机. 2,镜头. 3,光源及控制器. 4,主机.(含采集卡) 相机我们一般就看这些参数:(图示是海康相机) 快门 ...

海康工业相机镜头阴影矫正LSC