matlab yolov2对整个测试集的图片进行预测的代码

最新推荐文章于 2025-04-06 14:10:22 发布

学习滋味有点甜

最新推荐文章于 2025-04-06 14:10:22 发布

阅读量1.7k

点赞数 2

文章标签：测试代码 yolo Matlab 目标检测

本文链接：https://blog.csdn.net/weixin_42899305/article/details/95198622

版权

这个是matlab代码对yolov2生成的模型进行测试的代码，他会在图片生成boundingbox.

close all;clear;clc
gpuDevice(1);

load standard_cell2_detector_yolov2_0611_epoch50_changeanchor.mat;
test_image_folder = './predictdata_900_900_crop';

test_output_folder = 'test_cell_detection;
#带有boundingbox的图片测试结果的输出路径
bbox_output_folder = 'cell2_yolov2_bbox_black_white'
test_files = dir(test_image_folder);
length(test_files)

%make output folder
if ~isdir(test_output_folder)
    fprintf(1, 'Making directory %s\n', test_output_folder);
    mkdir(test_output_folder);
end

if ~isdir(bbox_output_folder)
    fprintf(1, 'Making directory %s\n', test_output_folder);
    mkdir(bbox_output_folder);
end

for k = 3 : length(test_files)
%for k = 3 : 10
    baseFileName = test_files(k).name;
    fullFileName = fullfile(test_image_folder, baseFileName);
    tic;
    I = imread(fullFileName);
    [bboxes,scores,labels] = detect(detector,I,'T